{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 3.0, "eval_steps": 500, "global_step": 3942, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0007614696363982486, "grad_norm": 1301.5599365234375, "learning_rate": 0.0, "loss": 1.7023, "step": 1 }, { "epoch": 0.0015229392727964973, "grad_norm": 1499.36865234375, "learning_rate": 2.0000000000000002e-07, "loss": 1.7043, "step": 2 }, { "epoch": 0.002284408909194746, "grad_norm": 1399.9569091796875, "learning_rate": 4.0000000000000003e-07, "loss": 1.6019, "step": 3 }, { "epoch": 0.0030458785455929945, "grad_norm": 1188.724365234375, "learning_rate": 6.000000000000001e-07, "loss": 1.7423, "step": 4 }, { "epoch": 0.003807348181991243, "grad_norm": 969.7083129882812, "learning_rate": 8.000000000000001e-07, "loss": 1.6516, "step": 5 }, { "epoch": 0.004568817818389492, "grad_norm": 962.1920776367188, "learning_rate": 1.0000000000000002e-06, "loss": 1.2925, "step": 6 }, { "epoch": 0.00533028745478774, "grad_norm": 1861.9527587890625, "learning_rate": 1.2000000000000002e-06, "loss": 1.4859, "step": 7 }, { "epoch": 0.006091757091185989, "grad_norm": 1748.2479248046875, "learning_rate": 1.4000000000000001e-06, "loss": 1.8132, "step": 8 }, { "epoch": 0.006853226727584237, "grad_norm": 1467.362060546875, "learning_rate": 1.6000000000000001e-06, "loss": 1.5848, "step": 9 }, { "epoch": 0.007614696363982486, "grad_norm": 1578.7880859375, "learning_rate": 1.8000000000000001e-06, "loss": 1.6334, "step": 10 }, { "epoch": 0.008376166000380734, "grad_norm": 1163.3890380859375, "learning_rate": 2.0000000000000003e-06, "loss": 1.4836, "step": 11 }, { "epoch": 0.009137635636778984, "grad_norm": 1121.4964599609375, "learning_rate": 2.2e-06, "loss": 1.7535, "step": 12 }, { "epoch": 0.009899105273177233, "grad_norm": 1086.3167724609375, "learning_rate": 2.4000000000000003e-06, "loss": 1.56, "step": 13 }, { "epoch": 0.01066057490957548, "grad_norm": 876.811767578125, "learning_rate": 2.6e-06, "loss": 1.2632, "step": 14 }, { "epoch": 0.011422044545973729, "grad_norm": 1044.1190185546875, "learning_rate": 2.8000000000000003e-06, "loss": 1.3597, "step": 15 }, { "epoch": 0.012183514182371978, "grad_norm": 1221.429443359375, "learning_rate": 3e-06, "loss": 1.7357, "step": 16 }, { "epoch": 0.012944983818770227, "grad_norm": 1007.3869018554688, "learning_rate": 3.2000000000000003e-06, "loss": 1.5886, "step": 17 }, { "epoch": 0.013706453455168474, "grad_norm": 1217.14404296875, "learning_rate": 3.4000000000000005e-06, "loss": 1.386, "step": 18 }, { "epoch": 0.014467923091566724, "grad_norm": 1351.4454345703125, "learning_rate": 3.6000000000000003e-06, "loss": 1.6928, "step": 19 }, { "epoch": 0.015229392727964973, "grad_norm": 1504.17041015625, "learning_rate": 3.8000000000000005e-06, "loss": 1.4281, "step": 20 }, { "epoch": 0.01599086236436322, "grad_norm": 1153.16162109375, "learning_rate": 4.000000000000001e-06, "loss": 1.1809, "step": 21 }, { "epoch": 0.01675233200076147, "grad_norm": 1124.1148681640625, "learning_rate": 4.2000000000000004e-06, "loss": 1.7594, "step": 22 }, { "epoch": 0.01751380163715972, "grad_norm": 1091.021484375, "learning_rate": 4.4e-06, "loss": 1.4819, "step": 23 }, { "epoch": 0.018275271273557967, "grad_norm": 802.41259765625, "learning_rate": 4.600000000000001e-06, "loss": 1.3459, "step": 24 }, { "epoch": 0.019036740909956214, "grad_norm": 1319.851318359375, "learning_rate": 4.800000000000001e-06, "loss": 1.6257, "step": 25 }, { "epoch": 0.019798210546354465, "grad_norm": 977.6575927734375, "learning_rate": 5e-06, "loss": 1.5093, "step": 26 }, { "epoch": 0.020559680182752713, "grad_norm": 774.690673828125, "learning_rate": 5.2e-06, "loss": 1.4782, "step": 27 }, { "epoch": 0.02132114981915096, "grad_norm": 937.8590087890625, "learning_rate": 5.400000000000001e-06, "loss": 1.6011, "step": 28 }, { "epoch": 0.02208261945554921, "grad_norm": 756.9816284179688, "learning_rate": 5.600000000000001e-06, "loss": 1.6296, "step": 29 }, { "epoch": 0.022844089091947458, "grad_norm": 1100.748291015625, "learning_rate": 5.8e-06, "loss": 1.6768, "step": 30 }, { "epoch": 0.02360555872834571, "grad_norm": 976.7691650390625, "learning_rate": 6e-06, "loss": 1.423, "step": 31 }, { "epoch": 0.024367028364743956, "grad_norm": 871.838134765625, "learning_rate": 6.200000000000001e-06, "loss": 1.3862, "step": 32 }, { "epoch": 0.025128498001142203, "grad_norm": 807.175537109375, "learning_rate": 6.4000000000000006e-06, "loss": 1.5093, "step": 33 }, { "epoch": 0.025889967637540454, "grad_norm": 964.3636474609375, "learning_rate": 6.600000000000001e-06, "loss": 1.7866, "step": 34 }, { "epoch": 0.0266514372739387, "grad_norm": 953.1375122070312, "learning_rate": 6.800000000000001e-06, "loss": 1.4513, "step": 35 }, { "epoch": 0.02741290691033695, "grad_norm": 893.89111328125, "learning_rate": 7e-06, "loss": 1.8488, "step": 36 }, { "epoch": 0.0281743765467352, "grad_norm": 833.2611694335938, "learning_rate": 7.2000000000000005e-06, "loss": 1.3522, "step": 37 }, { "epoch": 0.028935846183133447, "grad_norm": 837.7489013671875, "learning_rate": 7.4e-06, "loss": 1.5152, "step": 38 }, { "epoch": 0.029697315819531698, "grad_norm": 814.172607421875, "learning_rate": 7.600000000000001e-06, "loss": 1.2972, "step": 39 }, { "epoch": 0.030458785455929945, "grad_norm": 810.1613159179688, "learning_rate": 7.800000000000002e-06, "loss": 1.5865, "step": 40 }, { "epoch": 0.031220255092328193, "grad_norm": 496.3629150390625, "learning_rate": 8.000000000000001e-06, "loss": 1.4072, "step": 41 }, { "epoch": 0.03198172472872644, "grad_norm": 620.7055053710938, "learning_rate": 8.2e-06, "loss": 1.732, "step": 42 }, { "epoch": 0.03274319436512469, "grad_norm": 1089.77294921875, "learning_rate": 8.400000000000001e-06, "loss": 1.4347, "step": 43 }, { "epoch": 0.03350466400152294, "grad_norm": 829.736328125, "learning_rate": 8.6e-06, "loss": 1.4516, "step": 44 }, { "epoch": 0.034266133637921185, "grad_norm": 1636.154052734375, "learning_rate": 8.8e-06, "loss": 1.8134, "step": 45 }, { "epoch": 0.03502760327431944, "grad_norm": 1268.646728515625, "learning_rate": 9e-06, "loss": 1.4501, "step": 46 }, { "epoch": 0.03578907291071769, "grad_norm": 590.24755859375, "learning_rate": 9.200000000000002e-06, "loss": 1.4093, "step": 47 }, { "epoch": 0.036550542547115934, "grad_norm": 687.4384765625, "learning_rate": 9.4e-06, "loss": 1.5449, "step": 48 }, { "epoch": 0.03731201218351418, "grad_norm": 532.1568603515625, "learning_rate": 9.600000000000001e-06, "loss": 1.463, "step": 49 }, { "epoch": 0.03807348181991243, "grad_norm": 606.5445556640625, "learning_rate": 9.800000000000001e-06, "loss": 1.7299, "step": 50 }, { "epoch": 0.038834951456310676, "grad_norm": 494.4815368652344, "learning_rate": 1e-05, "loss": 1.6656, "step": 51 }, { "epoch": 0.03959642109270893, "grad_norm": 478.6385192871094, "learning_rate": 9.99999628743664e-06, "loss": 1.6203, "step": 52 }, { "epoch": 0.04035789072910718, "grad_norm": 495.6545104980469, "learning_rate": 9.999985149752072e-06, "loss": 1.4333, "step": 53 }, { "epoch": 0.041119360365505425, "grad_norm": 489.4484558105469, "learning_rate": 9.999966586962836e-06, "loss": 1.3458, "step": 54 }, { "epoch": 0.04188083000190367, "grad_norm": 470.57647705078125, "learning_rate": 9.999940599096498e-06, "loss": 1.563, "step": 55 }, { "epoch": 0.04264229963830192, "grad_norm": 402.5634460449219, "learning_rate": 9.999907186191652e-06, "loss": 1.3983, "step": 56 }, { "epoch": 0.043403769274700174, "grad_norm": 365.5474548339844, "learning_rate": 9.999866348297913e-06, "loss": 1.5199, "step": 57 }, { "epoch": 0.04416523891109842, "grad_norm": 389.02984619140625, "learning_rate": 9.99981808547593e-06, "loss": 1.3679, "step": 58 }, { "epoch": 0.04492670854749667, "grad_norm": 384.76470947265625, "learning_rate": 9.999762397797375e-06, "loss": 1.3927, "step": 59 }, { "epoch": 0.045688178183894916, "grad_norm": 380.3616638183594, "learning_rate": 9.999699285344943e-06, "loss": 1.4385, "step": 60 }, { "epoch": 0.04644964782029316, "grad_norm": 451.0272216796875, "learning_rate": 9.99962874821236e-06, "loss": 1.7561, "step": 61 }, { "epoch": 0.04721111745669142, "grad_norm": 546.6317138671875, "learning_rate": 9.999550786504374e-06, "loss": 1.6178, "step": 62 }, { "epoch": 0.047972587093089665, "grad_norm": 302.81549072265625, "learning_rate": 9.99946540033676e-06, "loss": 1.2085, "step": 63 }, { "epoch": 0.04873405672948791, "grad_norm": 344.7528076171875, "learning_rate": 9.999372589836319e-06, "loss": 1.2769, "step": 64 }, { "epoch": 0.04949552636588616, "grad_norm": 418.7420654296875, "learning_rate": 9.999272355140878e-06, "loss": 1.4762, "step": 65 }, { "epoch": 0.05025699600228441, "grad_norm": 323.90228271484375, "learning_rate": 9.999164696399287e-06, "loss": 1.2773, "step": 66 }, { "epoch": 0.051018465638682654, "grad_norm": 257.0769958496094, "learning_rate": 9.999049613771421e-06, "loss": 1.2942, "step": 67 }, { "epoch": 0.05177993527508091, "grad_norm": 366.9392395019531, "learning_rate": 9.998927107428181e-06, "loss": 1.3762, "step": 68 }, { "epoch": 0.052541404911479156, "grad_norm": 319.0862731933594, "learning_rate": 9.998797177551494e-06, "loss": 1.4763, "step": 69 }, { "epoch": 0.0533028745478774, "grad_norm": 309.5851135253906, "learning_rate": 9.998659824334307e-06, "loss": 1.2058, "step": 70 }, { "epoch": 0.05406434418427565, "grad_norm": 331.6963195800781, "learning_rate": 9.998515047980595e-06, "loss": 1.4926, "step": 71 }, { "epoch": 0.0548258138206739, "grad_norm": 267.27130126953125, "learning_rate": 9.998362848705351e-06, "loss": 1.3501, "step": 72 }, { "epoch": 0.05558728345707215, "grad_norm": 342.3760681152344, "learning_rate": 9.9982032267346e-06, "loss": 1.4401, "step": 73 }, { "epoch": 0.0563487530934704, "grad_norm": 256.15966796875, "learning_rate": 9.99803618230538e-06, "loss": 1.0673, "step": 74 }, { "epoch": 0.05711022272986865, "grad_norm": 294.26556396484375, "learning_rate": 9.997861715665758e-06, "loss": 1.4545, "step": 75 }, { "epoch": 0.057871692366266894, "grad_norm": 309.1824035644531, "learning_rate": 9.997679827074822e-06, "loss": 1.3322, "step": 76 }, { "epoch": 0.05863316200266514, "grad_norm": 281.6103515625, "learning_rate": 9.997490516802682e-06, "loss": 1.1179, "step": 77 }, { "epoch": 0.059394631639063396, "grad_norm": 259.3569641113281, "learning_rate": 9.997293785130464e-06, "loss": 1.413, "step": 78 }, { "epoch": 0.06015610127546164, "grad_norm": 312.6566467285156, "learning_rate": 9.997089632350325e-06, "loss": 1.4359, "step": 79 }, { "epoch": 0.06091757091185989, "grad_norm": 202.5186767578125, "learning_rate": 9.996878058765434e-06, "loss": 1.2213, "step": 80 }, { "epoch": 0.06167904054825814, "grad_norm": 235.61985778808594, "learning_rate": 9.996659064689985e-06, "loss": 1.089, "step": 81 }, { "epoch": 0.062440510184656385, "grad_norm": 343.52587890625, "learning_rate": 9.996432650449186e-06, "loss": 1.5558, "step": 82 }, { "epoch": 0.06320197982105463, "grad_norm": 438.2514343261719, "learning_rate": 9.996198816379273e-06, "loss": 1.4458, "step": 83 }, { "epoch": 0.06396344945745289, "grad_norm": 333.7451477050781, "learning_rate": 9.99595756282749e-06, "loss": 1.2887, "step": 84 }, { "epoch": 0.06472491909385113, "grad_norm": 232.31222534179688, "learning_rate": 9.99570889015211e-06, "loss": 1.2663, "step": 85 }, { "epoch": 0.06548638873024938, "grad_norm": 215.1619415283203, "learning_rate": 9.995452798722416e-06, "loss": 1.0614, "step": 86 }, { "epoch": 0.06624785836664764, "grad_norm": 255.4922332763672, "learning_rate": 9.99518928891871e-06, "loss": 1.4337, "step": 87 }, { "epoch": 0.06700932800304588, "grad_norm": 550.9943237304688, "learning_rate": 9.99491836113231e-06, "loss": 1.48, "step": 88 }, { "epoch": 0.06777079763944413, "grad_norm": 288.2671203613281, "learning_rate": 9.994640015765552e-06, "loss": 1.2056, "step": 89 }, { "epoch": 0.06853226727584237, "grad_norm": 231.81983947753906, "learning_rate": 9.994354253231783e-06, "loss": 1.0956, "step": 90 }, { "epoch": 0.06929373691224062, "grad_norm": 417.0189514160156, "learning_rate": 9.994061073955375e-06, "loss": 1.0774, "step": 91 }, { "epoch": 0.07005520654863888, "grad_norm": 219.20413208007812, "learning_rate": 9.993760478371697e-06, "loss": 1.3335, "step": 92 }, { "epoch": 0.07081667618503712, "grad_norm": 282.10443115234375, "learning_rate": 9.99345246692715e-06, "loss": 1.3221, "step": 93 }, { "epoch": 0.07157814582143537, "grad_norm": 213.92098999023438, "learning_rate": 9.993137040079131e-06, "loss": 1.3243, "step": 94 }, { "epoch": 0.07233961545783361, "grad_norm": 158.23342895507812, "learning_rate": 9.992814198296061e-06, "loss": 1.1767, "step": 95 }, { "epoch": 0.07310108509423187, "grad_norm": 177.14779663085938, "learning_rate": 9.992483942057369e-06, "loss": 1.2758, "step": 96 }, { "epoch": 0.07386255473063012, "grad_norm": 254.54237365722656, "learning_rate": 9.992146271853491e-06, "loss": 1.2782, "step": 97 }, { "epoch": 0.07462402436702836, "grad_norm": 202.73121643066406, "learning_rate": 9.991801188185878e-06, "loss": 1.2328, "step": 98 }, { "epoch": 0.07538549400342662, "grad_norm": 198.03977966308594, "learning_rate": 9.991448691566989e-06, "loss": 0.9788, "step": 99 }, { "epoch": 0.07614696363982486, "grad_norm": 203.24838256835938, "learning_rate": 9.991088782520286e-06, "loss": 1.4558, "step": 100 }, { "epoch": 0.07690843327622311, "grad_norm": 160.49710083007812, "learning_rate": 9.990721461580247e-06, "loss": 0.9447, "step": 101 }, { "epoch": 0.07766990291262135, "grad_norm": 162.8723907470703, "learning_rate": 9.990346729292353e-06, "loss": 1.0596, "step": 102 }, { "epoch": 0.0784313725490196, "grad_norm": 218.8311767578125, "learning_rate": 9.989964586213087e-06, "loss": 1.4439, "step": 103 }, { "epoch": 0.07919284218541786, "grad_norm": 243.22412109375, "learning_rate": 9.989575032909946e-06, "loss": 1.2728, "step": 104 }, { "epoch": 0.0799543118218161, "grad_norm": 235.75650024414062, "learning_rate": 9.989178069961422e-06, "loss": 1.1584, "step": 105 }, { "epoch": 0.08071578145821436, "grad_norm": 214.09156799316406, "learning_rate": 9.98877369795702e-06, "loss": 1.1591, "step": 106 }, { "epoch": 0.0814772510946126, "grad_norm": 205.96751403808594, "learning_rate": 9.988361917497238e-06, "loss": 1.114, "step": 107 }, { "epoch": 0.08223872073101085, "grad_norm": 230.2100830078125, "learning_rate": 9.987942729193584e-06, "loss": 1.4029, "step": 108 }, { "epoch": 0.0830001903674091, "grad_norm": 246.6984405517578, "learning_rate": 9.987516133668561e-06, "loss": 1.5481, "step": 109 }, { "epoch": 0.08376166000380734, "grad_norm": 313.53570556640625, "learning_rate": 9.987082131555674e-06, "loss": 0.9881, "step": 110 }, { "epoch": 0.0845231296402056, "grad_norm": 190.1482391357422, "learning_rate": 9.98664072349943e-06, "loss": 1.0438, "step": 111 }, { "epoch": 0.08528459927660384, "grad_norm": 230.99050903320312, "learning_rate": 9.986191910155328e-06, "loss": 1.3698, "step": 112 }, { "epoch": 0.0860460689130021, "grad_norm": 224.8727569580078, "learning_rate": 9.985735692189869e-06, "loss": 1.21, "step": 113 }, { "epoch": 0.08680753854940035, "grad_norm": 193.0661163330078, "learning_rate": 9.985272070280545e-06, "loss": 1.3077, "step": 114 }, { "epoch": 0.08756900818579859, "grad_norm": 156.9528045654297, "learning_rate": 9.984801045115852e-06, "loss": 0.9841, "step": 115 }, { "epoch": 0.08833047782219684, "grad_norm": 245.15037536621094, "learning_rate": 9.98432261739527e-06, "loss": 1.2538, "step": 116 }, { "epoch": 0.08909194745859508, "grad_norm": 278.6109313964844, "learning_rate": 9.983836787829277e-06, "loss": 1.0621, "step": 117 }, { "epoch": 0.08985341709499334, "grad_norm": 313.7957763671875, "learning_rate": 9.983343557139344e-06, "loss": 1.4762, "step": 118 }, { "epoch": 0.09061488673139159, "grad_norm": 160.8702392578125, "learning_rate": 9.98284292605793e-06, "loss": 1.1753, "step": 119 }, { "epoch": 0.09137635636778983, "grad_norm": 222.89654541015625, "learning_rate": 9.982334895328483e-06, "loss": 1.4998, "step": 120 }, { "epoch": 0.09213782600418809, "grad_norm": 214.5303955078125, "learning_rate": 9.981819465705443e-06, "loss": 1.2822, "step": 121 }, { "epoch": 0.09289929564058633, "grad_norm": 175.49874877929688, "learning_rate": 9.981296637954238e-06, "loss": 1.2615, "step": 122 }, { "epoch": 0.09366076527698458, "grad_norm": 167.6100616455078, "learning_rate": 9.980766412851277e-06, "loss": 1.4586, "step": 123 }, { "epoch": 0.09442223491338284, "grad_norm": 125.47764587402344, "learning_rate": 9.98022879118396e-06, "loss": 0.9281, "step": 124 }, { "epoch": 0.09518370454978108, "grad_norm": 166.0843505859375, "learning_rate": 9.979683773750668e-06, "loss": 1.0866, "step": 125 }, { "epoch": 0.09594517418617933, "grad_norm": 155.88449096679688, "learning_rate": 9.979131361360766e-06, "loss": 1.1467, "step": 126 }, { "epoch": 0.09670664382257757, "grad_norm": 163.07427978515625, "learning_rate": 9.9785715548346e-06, "loss": 1.0676, "step": 127 }, { "epoch": 0.09746811345897582, "grad_norm": 156.6132049560547, "learning_rate": 9.978004355003498e-06, "loss": 1.2959, "step": 128 }, { "epoch": 0.09822958309537407, "grad_norm": 193.43089294433594, "learning_rate": 9.977429762709765e-06, "loss": 1.1902, "step": 129 }, { "epoch": 0.09899105273177232, "grad_norm": 142.8735809326172, "learning_rate": 9.976847778806684e-06, "loss": 1.0323, "step": 130 }, { "epoch": 0.09975252236817057, "grad_norm": 154.01123046875, "learning_rate": 9.976258404158518e-06, "loss": 0.9512, "step": 131 }, { "epoch": 0.10051399200456881, "grad_norm": 183.3456573486328, "learning_rate": 9.975661639640503e-06, "loss": 1.0477, "step": 132 }, { "epoch": 0.10127546164096707, "grad_norm": 191.70790100097656, "learning_rate": 9.975057486138849e-06, "loss": 1.2902, "step": 133 }, { "epoch": 0.10203693127736531, "grad_norm": 156.32797241210938, "learning_rate": 9.97444594455074e-06, "loss": 1.1891, "step": 134 }, { "epoch": 0.10279840091376356, "grad_norm": 117.63006591796875, "learning_rate": 9.973827015784328e-06, "loss": 0.8506, "step": 135 }, { "epoch": 0.10355987055016182, "grad_norm": 125.6243667602539, "learning_rate": 9.973200700758741e-06, "loss": 1.1157, "step": 136 }, { "epoch": 0.10432134018656006, "grad_norm": 143.13687133789062, "learning_rate": 9.972567000404072e-06, "loss": 1.3107, "step": 137 }, { "epoch": 0.10508280982295831, "grad_norm": 136.7623291015625, "learning_rate": 9.97192591566138e-06, "loss": 0.7335, "step": 138 }, { "epoch": 0.10584427945935655, "grad_norm": 169.33311462402344, "learning_rate": 9.971277447482696e-06, "loss": 1.3181, "step": 139 }, { "epoch": 0.1066057490957548, "grad_norm": 143.11961364746094, "learning_rate": 9.970621596831007e-06, "loss": 0.6787, "step": 140 }, { "epoch": 0.10736721873215306, "grad_norm": 122.05813598632812, "learning_rate": 9.969958364680272e-06, "loss": 1.0809, "step": 141 }, { "epoch": 0.1081286883685513, "grad_norm": 120.5428695678711, "learning_rate": 9.969287752015406e-06, "loss": 1.0224, "step": 142 }, { "epoch": 0.10889015800494956, "grad_norm": 105.74716186523438, "learning_rate": 9.968609759832284e-06, "loss": 0.9593, "step": 143 }, { "epoch": 0.1096516276413478, "grad_norm": 199.15542602539062, "learning_rate": 9.967924389137745e-06, "loss": 1.0812, "step": 144 }, { "epoch": 0.11041309727774605, "grad_norm": 143.60145568847656, "learning_rate": 9.967231640949579e-06, "loss": 1.0824, "step": 145 }, { "epoch": 0.1111745669141443, "grad_norm": 117.3849868774414, "learning_rate": 9.966531516296536e-06, "loss": 1.0903, "step": 146 }, { "epoch": 0.11193603655054254, "grad_norm": 136.92623901367188, "learning_rate": 9.965824016218317e-06, "loss": 1.1075, "step": 147 }, { "epoch": 0.1126975061869408, "grad_norm": 125.6374740600586, "learning_rate": 9.96510914176558e-06, "loss": 0.9472, "step": 148 }, { "epoch": 0.11345897582333904, "grad_norm": 116.42830657958984, "learning_rate": 9.96438689399993e-06, "loss": 1.1378, "step": 149 }, { "epoch": 0.1142204454597373, "grad_norm": 99.17476654052734, "learning_rate": 9.963657273993928e-06, "loss": 0.9665, "step": 150 }, { "epoch": 0.11498191509613555, "grad_norm": 105.80054473876953, "learning_rate": 9.96292028283107e-06, "loss": 1.0024, "step": 151 }, { "epoch": 0.11574338473253379, "grad_norm": 114.62339782714844, "learning_rate": 9.962175921605812e-06, "loss": 1.1685, "step": 152 }, { "epoch": 0.11650485436893204, "grad_norm": 125.38658142089844, "learning_rate": 9.961424191423547e-06, "loss": 1.0197, "step": 153 }, { "epoch": 0.11726632400533028, "grad_norm": 129.48753356933594, "learning_rate": 9.960665093400616e-06, "loss": 1.1579, "step": 154 }, { "epoch": 0.11802779364172854, "grad_norm": 117.82019805908203, "learning_rate": 9.959898628664298e-06, "loss": 1.0146, "step": 155 }, { "epoch": 0.11878926327812679, "grad_norm": 151.73013305664062, "learning_rate": 9.95912479835281e-06, "loss": 1.3081, "step": 156 }, { "epoch": 0.11955073291452503, "grad_norm": 141.16104125976562, "learning_rate": 9.958343603615311e-06, "loss": 1.2722, "step": 157 }, { "epoch": 0.12031220255092329, "grad_norm": 94.05651092529297, "learning_rate": 9.957555045611897e-06, "loss": 0.8151, "step": 158 }, { "epoch": 0.12107367218732153, "grad_norm": 118.12832641601562, "learning_rate": 9.956759125513593e-06, "loss": 1.1914, "step": 159 }, { "epoch": 0.12183514182371978, "grad_norm": 155.52194213867188, "learning_rate": 9.955955844502365e-06, "loss": 1.2095, "step": 160 }, { "epoch": 0.12259661146011802, "grad_norm": 120.84684753417969, "learning_rate": 9.955145203771101e-06, "loss": 1.034, "step": 161 }, { "epoch": 0.12335808109651628, "grad_norm": 200.08302307128906, "learning_rate": 9.954327204523626e-06, "loss": 1.0109, "step": 162 }, { "epoch": 0.12411955073291453, "grad_norm": 91.24543762207031, "learning_rate": 9.953501847974689e-06, "loss": 0.889, "step": 163 }, { "epoch": 0.12488102036931277, "grad_norm": 122.83014678955078, "learning_rate": 9.952669135349963e-06, "loss": 0.9426, "step": 164 }, { "epoch": 0.125642490005711, "grad_norm": 116.14713287353516, "learning_rate": 9.951829067886053e-06, "loss": 0.9366, "step": 165 }, { "epoch": 0.12640395964210926, "grad_norm": 100.40796661376953, "learning_rate": 9.950981646830476e-06, "loss": 0.9302, "step": 166 }, { "epoch": 0.12716542927850752, "grad_norm": 123.98023223876953, "learning_rate": 9.950126873441674e-06, "loss": 1.2609, "step": 167 }, { "epoch": 0.12792689891490577, "grad_norm": 90.9245834350586, "learning_rate": 9.949264748989009e-06, "loss": 0.6413, "step": 168 }, { "epoch": 0.12868836855130403, "grad_norm": 121.35209655761719, "learning_rate": 9.948395274752754e-06, "loss": 0.9599, "step": 169 }, { "epoch": 0.12944983818770225, "grad_norm": 95.39146423339844, "learning_rate": 9.947518452024105e-06, "loss": 1.0403, "step": 170 }, { "epoch": 0.1302113078241005, "grad_norm": 112.46849060058594, "learning_rate": 9.946634282105164e-06, "loss": 1.1076, "step": 171 }, { "epoch": 0.13097277746049876, "grad_norm": 119.87206268310547, "learning_rate": 9.945742766308946e-06, "loss": 0.7471, "step": 172 }, { "epoch": 0.13173424709689702, "grad_norm": 159.3794708251953, "learning_rate": 9.944843905959372e-06, "loss": 1.0218, "step": 173 }, { "epoch": 0.13249571673329527, "grad_norm": 143.2695770263672, "learning_rate": 9.943937702391275e-06, "loss": 1.1849, "step": 174 }, { "epoch": 0.1332571863696935, "grad_norm": 129.3036346435547, "learning_rate": 9.94302415695039e-06, "loss": 1.0408, "step": 175 }, { "epoch": 0.13401865600609175, "grad_norm": 112.15624237060547, "learning_rate": 9.942103270993356e-06, "loss": 0.9284, "step": 176 }, { "epoch": 0.13478012564249, "grad_norm": 351.1947937011719, "learning_rate": 9.94117504588771e-06, "loss": 1.0657, "step": 177 }, { "epoch": 0.13554159527888826, "grad_norm": 198.4166259765625, "learning_rate": 9.940239483011892e-06, "loss": 0.8037, "step": 178 }, { "epoch": 0.13630306491528651, "grad_norm": 111.71124267578125, "learning_rate": 9.939296583755233e-06, "loss": 0.9574, "step": 179 }, { "epoch": 0.13706453455168474, "grad_norm": 103.34873962402344, "learning_rate": 9.938346349517966e-06, "loss": 1.1653, "step": 180 }, { "epoch": 0.137826004188083, "grad_norm": 118.38048553466797, "learning_rate": 9.937388781711213e-06, "loss": 1.0831, "step": 181 }, { "epoch": 0.13858747382448125, "grad_norm": 362.5590515136719, "learning_rate": 9.936423881756983e-06, "loss": 1.0866, "step": 182 }, { "epoch": 0.1393489434608795, "grad_norm": 262.0923156738281, "learning_rate": 9.93545165108818e-06, "loss": 0.896, "step": 183 }, { "epoch": 0.14011041309727776, "grad_norm": 146.92994689941406, "learning_rate": 9.934472091148588e-06, "loss": 1.0639, "step": 184 }, { "epoch": 0.14087188273367598, "grad_norm": 94.61048126220703, "learning_rate": 9.933485203392881e-06, "loss": 0.9745, "step": 185 }, { "epoch": 0.14163335237007424, "grad_norm": 109.75193786621094, "learning_rate": 9.932490989286614e-06, "loss": 1.0192, "step": 186 }, { "epoch": 0.1423948220064725, "grad_norm": 100.21412658691406, "learning_rate": 9.931489450306216e-06, "loss": 1.1705, "step": 187 }, { "epoch": 0.14315629164287075, "grad_norm": 103.8467788696289, "learning_rate": 9.930480587939e-06, "loss": 0.8971, "step": 188 }, { "epoch": 0.143917761279269, "grad_norm": 138.01211547851562, "learning_rate": 9.929464403683152e-06, "loss": 0.9786, "step": 189 }, { "epoch": 0.14467923091566723, "grad_norm": 94.98514556884766, "learning_rate": 9.92844089904773e-06, "loss": 0.908, "step": 190 }, { "epoch": 0.14544070055206548, "grad_norm": 77.64883422851562, "learning_rate": 9.927410075552666e-06, "loss": 0.8565, "step": 191 }, { "epoch": 0.14620217018846374, "grad_norm": 87.6455078125, "learning_rate": 9.92637193472876e-06, "loss": 0.978, "step": 192 }, { "epoch": 0.146963639824862, "grad_norm": 75.11997985839844, "learning_rate": 9.925326478117673e-06, "loss": 0.699, "step": 193 }, { "epoch": 0.14772510946126025, "grad_norm": 83.82418823242188, "learning_rate": 9.92427370727194e-06, "loss": 0.8723, "step": 194 }, { "epoch": 0.14848657909765847, "grad_norm": 118.23309326171875, "learning_rate": 9.92321362375495e-06, "loss": 1.1355, "step": 195 }, { "epoch": 0.14924804873405673, "grad_norm": 80.46004486083984, "learning_rate": 9.922146229140952e-06, "loss": 0.9206, "step": 196 }, { "epoch": 0.15000951837045498, "grad_norm": 114.65190887451172, "learning_rate": 9.921071525015057e-06, "loss": 1.0244, "step": 197 }, { "epoch": 0.15077098800685323, "grad_norm": 98.35903930664062, "learning_rate": 9.919989512973227e-06, "loss": 1.2852, "step": 198 }, { "epoch": 0.1515324576432515, "grad_norm": 97.59054565429688, "learning_rate": 9.918900194622277e-06, "loss": 0.9721, "step": 199 }, { "epoch": 0.15229392727964972, "grad_norm": 109.65848541259766, "learning_rate": 9.91780357157987e-06, "loss": 1.1749, "step": 200 }, { "epoch": 0.15305539691604797, "grad_norm": 85.1619644165039, "learning_rate": 9.916699645474525e-06, "loss": 0.9321, "step": 201 }, { "epoch": 0.15381686655244622, "grad_norm": 95.05500030517578, "learning_rate": 9.915588417945595e-06, "loss": 0.9065, "step": 202 }, { "epoch": 0.15457833618884448, "grad_norm": 100.80270385742188, "learning_rate": 9.914469890643282e-06, "loss": 1.1461, "step": 203 }, { "epoch": 0.1553398058252427, "grad_norm": 92.37889862060547, "learning_rate": 9.913344065228629e-06, "loss": 0.9546, "step": 204 }, { "epoch": 0.15610127546164096, "grad_norm": 103.6626968383789, "learning_rate": 9.912210943373515e-06, "loss": 1.0895, "step": 205 }, { "epoch": 0.1568627450980392, "grad_norm": 102.84590148925781, "learning_rate": 9.911070526760652e-06, "loss": 1.1842, "step": 206 }, { "epoch": 0.15762421473443747, "grad_norm": 85.6246566772461, "learning_rate": 9.90992281708359e-06, "loss": 0.8207, "step": 207 }, { "epoch": 0.15838568437083572, "grad_norm": 100.24021911621094, "learning_rate": 9.908767816046709e-06, "loss": 1.0015, "step": 208 }, { "epoch": 0.15914715400723395, "grad_norm": 61.58012008666992, "learning_rate": 9.90760552536521e-06, "loss": 0.8035, "step": 209 }, { "epoch": 0.1599086236436322, "grad_norm": 71.4411392211914, "learning_rate": 9.906435946765127e-06, "loss": 0.9014, "step": 210 }, { "epoch": 0.16067009328003046, "grad_norm": 115.7383041381836, "learning_rate": 9.905259081983314e-06, "loss": 1.0404, "step": 211 }, { "epoch": 0.1614315629164287, "grad_norm": 76.10044860839844, "learning_rate": 9.904074932767443e-06, "loss": 0.9077, "step": 212 }, { "epoch": 0.16219303255282697, "grad_norm": 86.42494201660156, "learning_rate": 9.902883500876008e-06, "loss": 1.0405, "step": 213 }, { "epoch": 0.1629545021892252, "grad_norm": 69.16837310791016, "learning_rate": 9.901684788078314e-06, "loss": 0.7598, "step": 214 }, { "epoch": 0.16371597182562345, "grad_norm": 72.36968994140625, "learning_rate": 9.90047879615448e-06, "loss": 0.9537, "step": 215 }, { "epoch": 0.1644774414620217, "grad_norm": 63.72938919067383, "learning_rate": 9.899265526895437e-06, "loss": 0.6775, "step": 216 }, { "epoch": 0.16523891109841995, "grad_norm": 112.95844268798828, "learning_rate": 9.898044982102916e-06, "loss": 1.349, "step": 217 }, { "epoch": 0.1660003807348182, "grad_norm": 112.42012786865234, "learning_rate": 9.896817163589462e-06, "loss": 1.3499, "step": 218 }, { "epoch": 0.16676185037121644, "grad_norm": 58.411903381347656, "learning_rate": 9.895582073178412e-06, "loss": 0.7196, "step": 219 }, { "epoch": 0.1675233200076147, "grad_norm": 78.74752044677734, "learning_rate": 9.894339712703911e-06, "loss": 1.0564, "step": 220 }, { "epoch": 0.16828478964401294, "grad_norm": 68.431884765625, "learning_rate": 9.89309008401089e-06, "loss": 0.8719, "step": 221 }, { "epoch": 0.1690462592804112, "grad_norm": 68.58363342285156, "learning_rate": 9.891833188955086e-06, "loss": 0.9706, "step": 222 }, { "epoch": 0.16980772891680945, "grad_norm": 73.43843841552734, "learning_rate": 9.890569029403016e-06, "loss": 0.8411, "step": 223 }, { "epoch": 0.17056919855320768, "grad_norm": 79.31640625, "learning_rate": 9.88929760723199e-06, "loss": 0.9633, "step": 224 }, { "epoch": 0.17133066818960593, "grad_norm": 66.7998275756836, "learning_rate": 9.888018924330101e-06, "loss": 0.8297, "step": 225 }, { "epoch": 0.1720921378260042, "grad_norm": 110.90506744384766, "learning_rate": 9.886732982596226e-06, "loss": 0.9995, "step": 226 }, { "epoch": 0.17285360746240244, "grad_norm": 66.6268310546875, "learning_rate": 9.885439783940023e-06, "loss": 0.83, "step": 227 }, { "epoch": 0.1736150770988007, "grad_norm": 76.32046508789062, "learning_rate": 9.884139330281921e-06, "loss": 1.0362, "step": 228 }, { "epoch": 0.17437654673519892, "grad_norm": 76.37635040283203, "learning_rate": 9.882831623553133e-06, "loss": 1.087, "step": 229 }, { "epoch": 0.17513801637159718, "grad_norm": 57.767845153808594, "learning_rate": 9.881516665695628e-06, "loss": 0.8016, "step": 230 }, { "epoch": 0.17589948600799543, "grad_norm": 78.45353698730469, "learning_rate": 9.880194458662159e-06, "loss": 0.8224, "step": 231 }, { "epoch": 0.17666095564439369, "grad_norm": 73.52896118164062, "learning_rate": 9.878865004416234e-06, "loss": 0.8957, "step": 232 }, { "epoch": 0.17742242528079194, "grad_norm": 83.69878387451172, "learning_rate": 9.877528304932125e-06, "loss": 1.1008, "step": 233 }, { "epoch": 0.17818389491719017, "grad_norm": 86.45205688476562, "learning_rate": 9.876184362194868e-06, "loss": 0.8303, "step": 234 }, { "epoch": 0.17894536455358842, "grad_norm": 99.48762512207031, "learning_rate": 9.87483317820025e-06, "loss": 1.2673, "step": 235 }, { "epoch": 0.17970683418998668, "grad_norm": 73.71369171142578, "learning_rate": 9.873474754954813e-06, "loss": 1.0541, "step": 236 }, { "epoch": 0.18046830382638493, "grad_norm": 60.825538635253906, "learning_rate": 9.872109094475849e-06, "loss": 0.7781, "step": 237 }, { "epoch": 0.18122977346278318, "grad_norm": 93.30781555175781, "learning_rate": 9.870736198791401e-06, "loss": 1.0317, "step": 238 }, { "epoch": 0.1819912430991814, "grad_norm": 86.9531478881836, "learning_rate": 9.869356069940255e-06, "loss": 1.0528, "step": 239 }, { "epoch": 0.18275271273557966, "grad_norm": 67.83726501464844, "learning_rate": 9.867968709971932e-06, "loss": 0.8781, "step": 240 }, { "epoch": 0.18351418237197792, "grad_norm": 72.33470153808594, "learning_rate": 9.866574120946701e-06, "loss": 1.0262, "step": 241 }, { "epoch": 0.18427565200837617, "grad_norm": 72.22496795654297, "learning_rate": 9.865172304935563e-06, "loss": 1.0208, "step": 242 }, { "epoch": 0.18503712164477443, "grad_norm": 59.84956359863281, "learning_rate": 9.863763264020245e-06, "loss": 0.6551, "step": 243 }, { "epoch": 0.18579859128117265, "grad_norm": 65.48404693603516, "learning_rate": 9.862347000293215e-06, "loss": 0.7664, "step": 244 }, { "epoch": 0.1865600609175709, "grad_norm": 85.38325500488281, "learning_rate": 9.860923515857654e-06, "loss": 0.8044, "step": 245 }, { "epoch": 0.18732153055396916, "grad_norm": 72.45954132080078, "learning_rate": 9.859492812827477e-06, "loss": 0.986, "step": 246 }, { "epoch": 0.18808300019036742, "grad_norm": 75.24809265136719, "learning_rate": 9.858054893327313e-06, "loss": 0.8028, "step": 247 }, { "epoch": 0.18884446982676567, "grad_norm": 79.74099731445312, "learning_rate": 9.85660975949251e-06, "loss": 1.0041, "step": 248 }, { "epoch": 0.1896059394631639, "grad_norm": 66.91632843017578, "learning_rate": 9.855157413469127e-06, "loss": 0.851, "step": 249 }, { "epoch": 0.19036740909956215, "grad_norm": 66.16661834716797, "learning_rate": 9.853697857413933e-06, "loss": 0.9538, "step": 250 }, { "epoch": 0.1911288787359604, "grad_norm": 70.9784927368164, "learning_rate": 9.85223109349441e-06, "loss": 1.0546, "step": 251 }, { "epoch": 0.19189034837235866, "grad_norm": 80.85004425048828, "learning_rate": 9.850757123888736e-06, "loss": 1.0673, "step": 252 }, { "epoch": 0.19265181800875691, "grad_norm": 70.5213394165039, "learning_rate": 9.849275950785795e-06, "loss": 0.7249, "step": 253 }, { "epoch": 0.19341328764515514, "grad_norm": 70.91667938232422, "learning_rate": 9.847787576385165e-06, "loss": 1.0218, "step": 254 }, { "epoch": 0.1941747572815534, "grad_norm": 80.40878295898438, "learning_rate": 9.846292002897123e-06, "loss": 0.885, "step": 255 }, { "epoch": 0.19493622691795165, "grad_norm": 64.33355712890625, "learning_rate": 9.844789232542629e-06, "loss": 0.8721, "step": 256 }, { "epoch": 0.1956976965543499, "grad_norm": 64.21064758300781, "learning_rate": 9.843279267553338e-06, "loss": 0.8722, "step": 257 }, { "epoch": 0.19645916619074813, "grad_norm": 228.57568359375, "learning_rate": 9.841762110171584e-06, "loss": 1.4434, "step": 258 }, { "epoch": 0.19722063582714638, "grad_norm": 56.00699234008789, "learning_rate": 9.840237762650389e-06, "loss": 0.8607, "step": 259 }, { "epoch": 0.19798210546354464, "grad_norm": 50.16679763793945, "learning_rate": 9.838706227253443e-06, "loss": 0.749, "step": 260 }, { "epoch": 0.1987435750999429, "grad_norm": 80.30974578857422, "learning_rate": 9.837167506255115e-06, "loss": 0.9376, "step": 261 }, { "epoch": 0.19950504473634115, "grad_norm": 67.49705505371094, "learning_rate": 9.835621601940449e-06, "loss": 0.996, "step": 262 }, { "epoch": 0.20026651437273937, "grad_norm": 87.47010040283203, "learning_rate": 9.834068516605146e-06, "loss": 1.2215, "step": 263 }, { "epoch": 0.20102798400913763, "grad_norm": 76.85598754882812, "learning_rate": 9.832508252555582e-06, "loss": 0.9047, "step": 264 }, { "epoch": 0.20178945364553588, "grad_norm": 72.64045715332031, "learning_rate": 9.830940812108787e-06, "loss": 0.9827, "step": 265 }, { "epoch": 0.20255092328193414, "grad_norm": 59.62710189819336, "learning_rate": 9.829366197592449e-06, "loss": 0.8007, "step": 266 }, { "epoch": 0.2033123929183324, "grad_norm": 71.46318817138672, "learning_rate": 9.827784411344912e-06, "loss": 0.8264, "step": 267 }, { "epoch": 0.20407386255473062, "grad_norm": 91.32540130615234, "learning_rate": 9.826195455715166e-06, "loss": 0.8318, "step": 268 }, { "epoch": 0.20483533219112887, "grad_norm": 107.7885513305664, "learning_rate": 9.824599333062855e-06, "loss": 1.2069, "step": 269 }, { "epoch": 0.20559680182752713, "grad_norm": 78.61428833007812, "learning_rate": 9.822996045758256e-06, "loss": 0.84, "step": 270 }, { "epoch": 0.20635827146392538, "grad_norm": 83.16557312011719, "learning_rate": 9.821385596182296e-06, "loss": 1.0659, "step": 271 }, { "epoch": 0.20711974110032363, "grad_norm": 69.2863540649414, "learning_rate": 9.819767986726529e-06, "loss": 0.802, "step": 272 }, { "epoch": 0.20788121073672186, "grad_norm": 72.9874038696289, "learning_rate": 9.818143219793152e-06, "loss": 1.0752, "step": 273 }, { "epoch": 0.20864268037312012, "grad_norm": 59.5941162109375, "learning_rate": 9.81651129779498e-06, "loss": 0.7396, "step": 274 }, { "epoch": 0.20940415000951837, "grad_norm": 63.05308151245117, "learning_rate": 9.81487222315546e-06, "loss": 0.8281, "step": 275 }, { "epoch": 0.21016561964591662, "grad_norm": 72.18460083007812, "learning_rate": 9.813225998308659e-06, "loss": 0.8861, "step": 276 }, { "epoch": 0.21092708928231488, "grad_norm": 66.82262420654297, "learning_rate": 9.811572625699263e-06, "loss": 0.7964, "step": 277 }, { "epoch": 0.2116885589187131, "grad_norm": 64.8218002319336, "learning_rate": 9.809912107782573e-06, "loss": 1.0174, "step": 278 }, { "epoch": 0.21245002855511136, "grad_norm": 73.7430648803711, "learning_rate": 9.8082444470245e-06, "loss": 1.1713, "step": 279 }, { "epoch": 0.2132114981915096, "grad_norm": 67.32575988769531, "learning_rate": 9.806569645901561e-06, "loss": 1.0453, "step": 280 }, { "epoch": 0.21397296782790787, "grad_norm": 58.80308532714844, "learning_rate": 9.804887706900881e-06, "loss": 0.8187, "step": 281 }, { "epoch": 0.21473443746430612, "grad_norm": 100.3160171508789, "learning_rate": 9.803198632520179e-06, "loss": 0.9593, "step": 282 }, { "epoch": 0.21549590710070435, "grad_norm": 64.89568328857422, "learning_rate": 9.801502425267773e-06, "loss": 0.7491, "step": 283 }, { "epoch": 0.2162573767371026, "grad_norm": 67.52566528320312, "learning_rate": 9.799799087662577e-06, "loss": 0.8042, "step": 284 }, { "epoch": 0.21701884637350086, "grad_norm": 51.88862228393555, "learning_rate": 9.798088622234088e-06, "loss": 0.8807, "step": 285 }, { "epoch": 0.2177803160098991, "grad_norm": 70.29974365234375, "learning_rate": 9.796371031522391e-06, "loss": 1.0593, "step": 286 }, { "epoch": 0.21854178564629737, "grad_norm": 60.2655143737793, "learning_rate": 9.794646318078153e-06, "loss": 0.8488, "step": 287 }, { "epoch": 0.2193032552826956, "grad_norm": 78.09272766113281, "learning_rate": 9.792914484462618e-06, "loss": 0.9485, "step": 288 }, { "epoch": 0.22006472491909385, "grad_norm": 55.70231628417969, "learning_rate": 9.791175533247597e-06, "loss": 0.8375, "step": 289 }, { "epoch": 0.2208261945554921, "grad_norm": 47.878211975097656, "learning_rate": 9.789429467015484e-06, "loss": 0.8201, "step": 290 }, { "epoch": 0.22158766419189035, "grad_norm": 59.19154357910156, "learning_rate": 9.787676288359225e-06, "loss": 0.9049, "step": 291 }, { "epoch": 0.2223491338282886, "grad_norm": 79.76387786865234, "learning_rate": 9.78591599988234e-06, "loss": 0.8522, "step": 292 }, { "epoch": 0.22311060346468684, "grad_norm": 58.5650749206543, "learning_rate": 9.7841486041989e-06, "loss": 0.8223, "step": 293 }, { "epoch": 0.2238720731010851, "grad_norm": 54.914093017578125, "learning_rate": 9.782374103933531e-06, "loss": 0.7369, "step": 294 }, { "epoch": 0.22463354273748334, "grad_norm": 56.13349151611328, "learning_rate": 9.780592501721414e-06, "loss": 0.9265, "step": 295 }, { "epoch": 0.2253950123738816, "grad_norm": 52.732635498046875, "learning_rate": 9.77880380020827e-06, "loss": 0.7705, "step": 296 }, { "epoch": 0.22615648201027985, "grad_norm": 54.85570526123047, "learning_rate": 9.777008002050367e-06, "loss": 1.1053, "step": 297 }, { "epoch": 0.22691795164667808, "grad_norm": 48.77083206176758, "learning_rate": 9.775205109914513e-06, "loss": 0.8562, "step": 298 }, { "epoch": 0.22767942128307633, "grad_norm": 48.549400329589844, "learning_rate": 9.773395126478047e-06, "loss": 0.7249, "step": 299 }, { "epoch": 0.2284408909194746, "grad_norm": 60.0245361328125, "learning_rate": 9.771578054428841e-06, "loss": 0.7139, "step": 300 }, { "epoch": 0.22920236055587284, "grad_norm": 55.8681755065918, "learning_rate": 9.769753896465291e-06, "loss": 0.7965, "step": 301 }, { "epoch": 0.2299638301922711, "grad_norm": 63.89664840698242, "learning_rate": 9.76792265529632e-06, "loss": 0.8464, "step": 302 }, { "epoch": 0.23072529982866932, "grad_norm": 71.57247161865234, "learning_rate": 9.766084333641366e-06, "loss": 1.0882, "step": 303 }, { "epoch": 0.23148676946506758, "grad_norm": 79.77525329589844, "learning_rate": 9.764238934230384e-06, "loss": 1.0725, "step": 304 }, { "epoch": 0.23224823910146583, "grad_norm": 67.99383544921875, "learning_rate": 9.76238645980384e-06, "loss": 0.8869, "step": 305 }, { "epoch": 0.23300970873786409, "grad_norm": 65.60368347167969, "learning_rate": 9.760526913112702e-06, "loss": 1.055, "step": 306 }, { "epoch": 0.23377117837426234, "grad_norm": 65.8590087890625, "learning_rate": 9.758660296918447e-06, "loss": 0.8672, "step": 307 }, { "epoch": 0.23453264801066057, "grad_norm": 59.94193649291992, "learning_rate": 9.756786613993048e-06, "loss": 0.8927, "step": 308 }, { "epoch": 0.23529411764705882, "grad_norm": 59.51659393310547, "learning_rate": 9.754905867118969e-06, "loss": 0.9401, "step": 309 }, { "epoch": 0.23605558728345707, "grad_norm": 58.14861297607422, "learning_rate": 9.753018059089168e-06, "loss": 0.795, "step": 310 }, { "epoch": 0.23681705691985533, "grad_norm": 54.271827697753906, "learning_rate": 9.751123192707089e-06, "loss": 0.8585, "step": 311 }, { "epoch": 0.23757852655625358, "grad_norm": 63.290191650390625, "learning_rate": 9.749221270786656e-06, "loss": 0.9526, "step": 312 }, { "epoch": 0.2383399961926518, "grad_norm": 48.89083480834961, "learning_rate": 9.747312296152268e-06, "loss": 0.8029, "step": 313 }, { "epoch": 0.23910146582905006, "grad_norm": 53.1309928894043, "learning_rate": 9.745396271638805e-06, "loss": 0.8306, "step": 314 }, { "epoch": 0.23986293546544832, "grad_norm": 53.48817443847656, "learning_rate": 9.74347320009161e-06, "loss": 0.8496, "step": 315 }, { "epoch": 0.24062440510184657, "grad_norm": 60.37258529663086, "learning_rate": 9.741543084366493e-06, "loss": 0.9068, "step": 316 }, { "epoch": 0.2413858747382448, "grad_norm": 68.24382781982422, "learning_rate": 9.739605927329725e-06, "loss": 1.0776, "step": 317 }, { "epoch": 0.24214734437464305, "grad_norm": 44.461395263671875, "learning_rate": 9.737661731858035e-06, "loss": 0.7072, "step": 318 }, { "epoch": 0.2429088140110413, "grad_norm": 50.15453338623047, "learning_rate": 9.7357105008386e-06, "loss": 0.7752, "step": 319 }, { "epoch": 0.24367028364743956, "grad_norm": 57.21116256713867, "learning_rate": 9.733752237169048e-06, "loss": 0.7657, "step": 320 }, { "epoch": 0.24443175328383782, "grad_norm": 53.94674301147461, "learning_rate": 9.731786943757451e-06, "loss": 0.8257, "step": 321 }, { "epoch": 0.24519322292023604, "grad_norm": 67.50530242919922, "learning_rate": 9.729814623522318e-06, "loss": 1.0285, "step": 322 }, { "epoch": 0.2459546925566343, "grad_norm": 61.62517166137695, "learning_rate": 9.727835279392597e-06, "loss": 0.8321, "step": 323 }, { "epoch": 0.24671616219303255, "grad_norm": 76.12077331542969, "learning_rate": 9.725848914307662e-06, "loss": 1.0304, "step": 324 }, { "epoch": 0.2474776318294308, "grad_norm": 58.61418151855469, "learning_rate": 9.723855531217316e-06, "loss": 0.9892, "step": 325 }, { "epoch": 0.24823910146582906, "grad_norm": 51.80055618286133, "learning_rate": 9.721855133081786e-06, "loss": 0.7943, "step": 326 }, { "epoch": 0.24900057110222729, "grad_norm": 50.69115447998047, "learning_rate": 9.71984772287171e-06, "loss": 0.8635, "step": 327 }, { "epoch": 0.24976204073862554, "grad_norm": 65.22866821289062, "learning_rate": 9.717833303568145e-06, "loss": 1.1264, "step": 328 }, { "epoch": 0.2505235103750238, "grad_norm": 64.84595489501953, "learning_rate": 9.715811878162556e-06, "loss": 1.1264, "step": 329 }, { "epoch": 0.251284980011422, "grad_norm": 47.65447235107422, "learning_rate": 9.713783449656807e-06, "loss": 0.823, "step": 330 }, { "epoch": 0.2520464496478203, "grad_norm": 46.5490608215332, "learning_rate": 9.71174802106317e-06, "loss": 0.8889, "step": 331 }, { "epoch": 0.25280791928421853, "grad_norm": 45.66133499145508, "learning_rate": 9.709705595404304e-06, "loss": 0.726, "step": 332 }, { "epoch": 0.2535693889206168, "grad_norm": 58.741180419921875, "learning_rate": 9.707656175713266e-06, "loss": 0.9001, "step": 333 }, { "epoch": 0.25433085855701504, "grad_norm": 56.66584396362305, "learning_rate": 9.705599765033498e-06, "loss": 0.8316, "step": 334 }, { "epoch": 0.25509232819341326, "grad_norm": 47.76698303222656, "learning_rate": 9.703536366418816e-06, "loss": 0.9308, "step": 335 }, { "epoch": 0.25585379782981155, "grad_norm": 52.65715789794922, "learning_rate": 9.701465982933423e-06, "loss": 0.9306, "step": 336 }, { "epoch": 0.2566152674662098, "grad_norm": 57.91140365600586, "learning_rate": 9.699388617651891e-06, "loss": 0.9152, "step": 337 }, { "epoch": 0.25737673710260806, "grad_norm": 46.12555694580078, "learning_rate": 9.697304273659161e-06, "loss": 0.8603, "step": 338 }, { "epoch": 0.2581382067390063, "grad_norm": 51.05317687988281, "learning_rate": 9.695212954050532e-06, "loss": 1.0602, "step": 339 }, { "epoch": 0.2588996763754045, "grad_norm": 45.56175994873047, "learning_rate": 9.693114661931673e-06, "loss": 0.7266, "step": 340 }, { "epoch": 0.2596611460118028, "grad_norm": 149.18678283691406, "learning_rate": 9.691009400418595e-06, "loss": 0.9806, "step": 341 }, { "epoch": 0.260422615648201, "grad_norm": 45.493690490722656, "learning_rate": 9.68889717263767e-06, "loss": 0.9679, "step": 342 }, { "epoch": 0.2611840852845993, "grad_norm": 66.79802703857422, "learning_rate": 9.686777981725606e-06, "loss": 0.9454, "step": 343 }, { "epoch": 0.2619455549209975, "grad_norm": 75.52727508544922, "learning_rate": 9.684651830829456e-06, "loss": 0.9213, "step": 344 }, { "epoch": 0.26270702455739575, "grad_norm": 54.42271423339844, "learning_rate": 9.682518723106608e-06, "loss": 0.9061, "step": 345 }, { "epoch": 0.26346849419379403, "grad_norm": 84.50782775878906, "learning_rate": 9.680378661724782e-06, "loss": 0.7799, "step": 346 }, { "epoch": 0.26422996383019226, "grad_norm": 56.192962646484375, "learning_rate": 9.678231649862024e-06, "loss": 0.7537, "step": 347 }, { "epoch": 0.26499143346659054, "grad_norm": 62.09931182861328, "learning_rate": 9.6760776907067e-06, "loss": 1.0033, "step": 348 }, { "epoch": 0.26575290310298877, "grad_norm": 48.45965576171875, "learning_rate": 9.673916787457492e-06, "loss": 0.845, "step": 349 }, { "epoch": 0.266514372739387, "grad_norm": 54.92905044555664, "learning_rate": 9.6717489433234e-06, "loss": 0.7935, "step": 350 }, { "epoch": 0.2672758423757853, "grad_norm": 44.289329528808594, "learning_rate": 9.669574161523723e-06, "loss": 0.7688, "step": 351 }, { "epoch": 0.2680373120121835, "grad_norm": 44.33442306518555, "learning_rate": 9.667392445288071e-06, "loss": 0.7193, "step": 352 }, { "epoch": 0.2687987816485818, "grad_norm": 51.590057373046875, "learning_rate": 9.665203797856346e-06, "loss": 0.7921, "step": 353 }, { "epoch": 0.26956025128498, "grad_norm": 45.111812591552734, "learning_rate": 9.663008222478747e-06, "loss": 0.6665, "step": 354 }, { "epoch": 0.27032172092137824, "grad_norm": 57.12602233886719, "learning_rate": 9.660805722415755e-06, "loss": 0.8825, "step": 355 }, { "epoch": 0.2710831905577765, "grad_norm": 50.83316421508789, "learning_rate": 9.658596300938143e-06, "loss": 0.881, "step": 356 }, { "epoch": 0.27184466019417475, "grad_norm": 52.06378936767578, "learning_rate": 9.656379961326954e-06, "loss": 0.9237, "step": 357 }, { "epoch": 0.27260612983057303, "grad_norm": 46.128719329833984, "learning_rate": 9.65415670687351e-06, "loss": 0.7108, "step": 358 }, { "epoch": 0.27336759946697126, "grad_norm": 40.89448547363281, "learning_rate": 9.651926540879401e-06, "loss": 0.7963, "step": 359 }, { "epoch": 0.2741290691033695, "grad_norm": 43.682315826416016, "learning_rate": 9.64968946665648e-06, "loss": 0.7105, "step": 360 }, { "epoch": 0.27489053873976776, "grad_norm": 45.00116729736328, "learning_rate": 9.647445487526857e-06, "loss": 0.676, "step": 361 }, { "epoch": 0.275652008376166, "grad_norm": 60.66935348510742, "learning_rate": 9.645194606822902e-06, "loss": 0.9782, "step": 362 }, { "epoch": 0.2764134780125643, "grad_norm": 62.74099349975586, "learning_rate": 9.642936827887223e-06, "loss": 0.8989, "step": 363 }, { "epoch": 0.2771749476489625, "grad_norm": 49.292667388916016, "learning_rate": 9.640672154072685e-06, "loss": 0.8945, "step": 364 }, { "epoch": 0.2779364172853607, "grad_norm": 45.407901763916016, "learning_rate": 9.638400588742385e-06, "loss": 0.7232, "step": 365 }, { "epoch": 0.278697886921759, "grad_norm": 92.1402816772461, "learning_rate": 9.636122135269655e-06, "loss": 0.6411, "step": 366 }, { "epoch": 0.27945935655815723, "grad_norm": 52.646549224853516, "learning_rate": 9.633836797038052e-06, "loss": 0.7967, "step": 367 }, { "epoch": 0.2802208261945555, "grad_norm": 51.09405517578125, "learning_rate": 9.631544577441367e-06, "loss": 0.8869, "step": 368 }, { "epoch": 0.28098229583095374, "grad_norm": 44.59373092651367, "learning_rate": 9.6292454798836e-06, "loss": 0.842, "step": 369 }, { "epoch": 0.28174376546735197, "grad_norm": 62.891151428222656, "learning_rate": 9.626939507778972e-06, "loss": 1.2426, "step": 370 }, { "epoch": 0.28250523510375025, "grad_norm": 47.443321228027344, "learning_rate": 9.624626664551908e-06, "loss": 0.8091, "step": 371 }, { "epoch": 0.2832667047401485, "grad_norm": 66.12074279785156, "learning_rate": 9.622306953637042e-06, "loss": 0.9351, "step": 372 }, { "epoch": 0.28402817437654676, "grad_norm": 63.217369079589844, "learning_rate": 9.6199803784792e-06, "loss": 0.8927, "step": 373 }, { "epoch": 0.284789644012945, "grad_norm": 49.39320373535156, "learning_rate": 9.617646942533403e-06, "loss": 0.7507, "step": 374 }, { "epoch": 0.2855511136493432, "grad_norm": 49.84722137451172, "learning_rate": 9.61530664926487e-06, "loss": 0.834, "step": 375 }, { "epoch": 0.2863125832857415, "grad_norm": 46.706214904785156, "learning_rate": 9.612959502148989e-06, "loss": 0.7922, "step": 376 }, { "epoch": 0.2870740529221397, "grad_norm": 46.903907775878906, "learning_rate": 9.610605504671338e-06, "loss": 0.9275, "step": 377 }, { "epoch": 0.287835522558538, "grad_norm": 42.682334899902344, "learning_rate": 9.608244660327657e-06, "loss": 0.8216, "step": 378 }, { "epoch": 0.28859699219493623, "grad_norm": 46.28004837036133, "learning_rate": 9.605876972623866e-06, "loss": 1.0026, "step": 379 }, { "epoch": 0.28935846183133446, "grad_norm": 46.75234603881836, "learning_rate": 9.603502445076038e-06, "loss": 0.7993, "step": 380 }, { "epoch": 0.29011993146773274, "grad_norm": 47.5782470703125, "learning_rate": 9.601121081210405e-06, "loss": 0.8264, "step": 381 }, { "epoch": 0.29088140110413097, "grad_norm": 66.08570861816406, "learning_rate": 9.598732884563357e-06, "loss": 1.1345, "step": 382 }, { "epoch": 0.29164287074052925, "grad_norm": 54.552268981933594, "learning_rate": 9.596337858681423e-06, "loss": 0.8592, "step": 383 }, { "epoch": 0.2924043403769275, "grad_norm": 44.23638916015625, "learning_rate": 9.593936007121279e-06, "loss": 0.7281, "step": 384 }, { "epoch": 0.2931658100133257, "grad_norm": 64.91778564453125, "learning_rate": 9.591527333449734e-06, "loss": 1.0959, "step": 385 }, { "epoch": 0.293927279649724, "grad_norm": 50.69198989868164, "learning_rate": 9.58911184124373e-06, "loss": 0.8607, "step": 386 }, { "epoch": 0.2946887492861222, "grad_norm": 44.16769790649414, "learning_rate": 9.586689534090336e-06, "loss": 0.8788, "step": 387 }, { "epoch": 0.2954502189225205, "grad_norm": 46.0577507019043, "learning_rate": 9.584260415586737e-06, "loss": 0.7654, "step": 388 }, { "epoch": 0.2962116885589187, "grad_norm": 53.36354446411133, "learning_rate": 9.581824489340235e-06, "loss": 0.6452, "step": 389 }, { "epoch": 0.29697315819531694, "grad_norm": 50.50668716430664, "learning_rate": 9.579381758968244e-06, "loss": 0.961, "step": 390 }, { "epoch": 0.2977346278317152, "grad_norm": 51.48319625854492, "learning_rate": 9.576932228098282e-06, "loss": 0.8882, "step": 391 }, { "epoch": 0.29849609746811345, "grad_norm": 45.250450134277344, "learning_rate": 9.574475900367962e-06, "loss": 0.7654, "step": 392 }, { "epoch": 0.29925756710451173, "grad_norm": 52.613136291503906, "learning_rate": 9.572012779424994e-06, "loss": 0.8605, "step": 393 }, { "epoch": 0.30001903674090996, "grad_norm": 53.87302017211914, "learning_rate": 9.569542868927174e-06, "loss": 0.8858, "step": 394 }, { "epoch": 0.3007805063773082, "grad_norm": 48.812095642089844, "learning_rate": 9.56706617254238e-06, "loss": 0.7843, "step": 395 }, { "epoch": 0.30154197601370647, "grad_norm": 45.33589553833008, "learning_rate": 9.564582693948575e-06, "loss": 0.8224, "step": 396 }, { "epoch": 0.3023034456501047, "grad_norm": 45.24856185913086, "learning_rate": 9.56209243683378e-06, "loss": 0.7123, "step": 397 }, { "epoch": 0.303064915286503, "grad_norm": 50.84047317504883, "learning_rate": 9.559595404896096e-06, "loss": 0.8103, "step": 398 }, { "epoch": 0.3038263849229012, "grad_norm": 52.10234832763672, "learning_rate": 9.557091601843676e-06, "loss": 0.9453, "step": 399 }, { "epoch": 0.30458785455929943, "grad_norm": 44.31291961669922, "learning_rate": 9.554581031394732e-06, "loss": 0.89, "step": 400 }, { "epoch": 0.3053493241956977, "grad_norm": 58.56898880004883, "learning_rate": 9.552063697277522e-06, "loss": 0.7098, "step": 401 }, { "epoch": 0.30611079383209594, "grad_norm": 66.1932373046875, "learning_rate": 9.549539603230355e-06, "loss": 1.0265, "step": 402 }, { "epoch": 0.30687226346849417, "grad_norm": 41.02724075317383, "learning_rate": 9.547008753001572e-06, "loss": 0.8262, "step": 403 }, { "epoch": 0.30763373310489245, "grad_norm": 43.16440200805664, "learning_rate": 9.544471150349551e-06, "loss": 0.8475, "step": 404 }, { "epoch": 0.3083952027412907, "grad_norm": 77.460693359375, "learning_rate": 9.541926799042695e-06, "loss": 1.1474, "step": 405 }, { "epoch": 0.30915667237768896, "grad_norm": 53.00200271606445, "learning_rate": 9.539375702859432e-06, "loss": 0.6984, "step": 406 }, { "epoch": 0.3099181420140872, "grad_norm": 80.07589721679688, "learning_rate": 9.536817865588203e-06, "loss": 0.7818, "step": 407 }, { "epoch": 0.3106796116504854, "grad_norm": 61.049251556396484, "learning_rate": 9.53425329102746e-06, "loss": 0.9013, "step": 408 }, { "epoch": 0.3114410812868837, "grad_norm": 59.325626373291016, "learning_rate": 9.531681982985666e-06, "loss": 0.9, "step": 409 }, { "epoch": 0.3122025509232819, "grad_norm": 50.168907165527344, "learning_rate": 9.529103945281273e-06, "loss": 0.696, "step": 410 }, { "epoch": 0.3129640205596802, "grad_norm": 98.40620422363281, "learning_rate": 9.526519181742735e-06, "loss": 0.7809, "step": 411 }, { "epoch": 0.3137254901960784, "grad_norm": 39.94493103027344, "learning_rate": 9.523927696208491e-06, "loss": 0.8265, "step": 412 }, { "epoch": 0.31448695983247665, "grad_norm": 37.19593048095703, "learning_rate": 9.521329492526964e-06, "loss": 0.5947, "step": 413 }, { "epoch": 0.31524842946887494, "grad_norm": 52.14117431640625, "learning_rate": 9.518724574556551e-06, "loss": 0.9694, "step": 414 }, { "epoch": 0.31600989910527316, "grad_norm": 45.72605514526367, "learning_rate": 9.516112946165619e-06, "loss": 0.752, "step": 415 }, { "epoch": 0.31677136874167144, "grad_norm": 55.8684196472168, "learning_rate": 9.513494611232507e-06, "loss": 0.9199, "step": 416 }, { "epoch": 0.31753283837806967, "grad_norm": 51.279266357421875, "learning_rate": 9.510869573645506e-06, "loss": 0.8778, "step": 417 }, { "epoch": 0.3182943080144679, "grad_norm": 49.07886505126953, "learning_rate": 9.508237837302862e-06, "loss": 0.9251, "step": 418 }, { "epoch": 0.3190557776508662, "grad_norm": 47.7920036315918, "learning_rate": 9.505599406112774e-06, "loss": 0.86, "step": 419 }, { "epoch": 0.3198172472872644, "grad_norm": 46.17924880981445, "learning_rate": 9.502954283993377e-06, "loss": 0.874, "step": 420 }, { "epoch": 0.3205787169236627, "grad_norm": 43.47404861450195, "learning_rate": 9.500302474872742e-06, "loss": 0.7282, "step": 421 }, { "epoch": 0.3213401865600609, "grad_norm": 41.86683654785156, "learning_rate": 9.497643982688878e-06, "loss": 0.7389, "step": 422 }, { "epoch": 0.32210165619645914, "grad_norm": 47.54405212402344, "learning_rate": 9.49497881138971e-06, "loss": 0.8291, "step": 423 }, { "epoch": 0.3228631258328574, "grad_norm": 47.38523483276367, "learning_rate": 9.492306964933085e-06, "loss": 0.7763, "step": 424 }, { "epoch": 0.32362459546925565, "grad_norm": 41.70780944824219, "learning_rate": 9.489628447286762e-06, "loss": 0.7601, "step": 425 }, { "epoch": 0.32438606510565393, "grad_norm": 44.05202102661133, "learning_rate": 9.48694326242841e-06, "loss": 0.9465, "step": 426 }, { "epoch": 0.32514753474205216, "grad_norm": 48.388675689697266, "learning_rate": 9.484251414345595e-06, "loss": 0.8215, "step": 427 }, { "epoch": 0.3259090043784504, "grad_norm": 44.47600555419922, "learning_rate": 9.48155290703578e-06, "loss": 0.9731, "step": 428 }, { "epoch": 0.32667047401484867, "grad_norm": 44.480506896972656, "learning_rate": 9.478847744506318e-06, "loss": 0.9429, "step": 429 }, { "epoch": 0.3274319436512469, "grad_norm": 56.323486328125, "learning_rate": 9.47613593077444e-06, "loss": 1.0475, "step": 430 }, { "epoch": 0.3281934132876452, "grad_norm": 54.73971939086914, "learning_rate": 9.473417469867263e-06, "loss": 0.973, "step": 431 }, { "epoch": 0.3289548829240434, "grad_norm": 41.304107666015625, "learning_rate": 9.470692365821767e-06, "loss": 0.775, "step": 432 }, { "epoch": 0.32971635256044163, "grad_norm": 43.46769714355469, "learning_rate": 9.467960622684803e-06, "loss": 0.8275, "step": 433 }, { "epoch": 0.3304778221968399, "grad_norm": 39.45466232299805, "learning_rate": 9.465222244513074e-06, "loss": 0.973, "step": 434 }, { "epoch": 0.33123929183323814, "grad_norm": 41.127464294433594, "learning_rate": 9.462477235373148e-06, "loss": 0.9163, "step": 435 }, { "epoch": 0.3320007614696364, "grad_norm": 43.90326690673828, "learning_rate": 9.459725599341427e-06, "loss": 0.8066, "step": 436 }, { "epoch": 0.33276223110603464, "grad_norm": 36.721168518066406, "learning_rate": 9.456967340504164e-06, "loss": 0.768, "step": 437 }, { "epoch": 0.33352370074243287, "grad_norm": 40.407039642333984, "learning_rate": 9.454202462957442e-06, "loss": 0.79, "step": 438 }, { "epoch": 0.33428517037883115, "grad_norm": 36.014339447021484, "learning_rate": 9.451430970807174e-06, "loss": 0.7859, "step": 439 }, { "epoch": 0.3350466400152294, "grad_norm": 36.85624694824219, "learning_rate": 9.448652868169098e-06, "loss": 0.8155, "step": 440 }, { "epoch": 0.33580810965162766, "grad_norm": 46.37972640991211, "learning_rate": 9.445868159168763e-06, "loss": 0.9041, "step": 441 }, { "epoch": 0.3365695792880259, "grad_norm": 43.00399398803711, "learning_rate": 9.443076847941535e-06, "loss": 0.9208, "step": 442 }, { "epoch": 0.3373310489244241, "grad_norm": 46.653106689453125, "learning_rate": 9.440278938632583e-06, "loss": 0.8929, "step": 443 }, { "epoch": 0.3380925185608224, "grad_norm": 40.778255462646484, "learning_rate": 9.43747443539687e-06, "loss": 0.7525, "step": 444 }, { "epoch": 0.3388539881972206, "grad_norm": 31.992992401123047, "learning_rate": 9.434663342399157e-06, "loss": 0.7598, "step": 445 }, { "epoch": 0.3396154578336189, "grad_norm": 46.3204231262207, "learning_rate": 9.431845663813988e-06, "loss": 0.9856, "step": 446 }, { "epoch": 0.34037692747001713, "grad_norm": 59.671539306640625, "learning_rate": 9.429021403825685e-06, "loss": 1.079, "step": 447 }, { "epoch": 0.34113839710641536, "grad_norm": 65.77830505371094, "learning_rate": 9.426190566628349e-06, "loss": 0.8555, "step": 448 }, { "epoch": 0.34189986674281364, "grad_norm": 48.69049835205078, "learning_rate": 9.423353156425843e-06, "loss": 0.7956, "step": 449 }, { "epoch": 0.34266133637921187, "grad_norm": 44.31893539428711, "learning_rate": 9.420509177431793e-06, "loss": 0.828, "step": 450 }, { "epoch": 0.34342280601561015, "grad_norm": 36.260643005371094, "learning_rate": 9.417658633869579e-06, "loss": 0.8325, "step": 451 }, { "epoch": 0.3441842756520084, "grad_norm": 41.35783767700195, "learning_rate": 9.414801529972333e-06, "loss": 0.9002, "step": 452 }, { "epoch": 0.3449457452884066, "grad_norm": 43.58637619018555, "learning_rate": 9.411937869982925e-06, "loss": 0.8771, "step": 453 }, { "epoch": 0.3457072149248049, "grad_norm": 38.9900016784668, "learning_rate": 9.409067658153964e-06, "loss": 0.8082, "step": 454 }, { "epoch": 0.3464686845612031, "grad_norm": 48.0144157409668, "learning_rate": 9.406190898747786e-06, "loss": 0.9711, "step": 455 }, { "epoch": 0.3472301541976014, "grad_norm": 43.04066848754883, "learning_rate": 9.40330759603645e-06, "loss": 0.971, "step": 456 }, { "epoch": 0.3479916238339996, "grad_norm": 44.24787902832031, "learning_rate": 9.400417754301739e-06, "loss": 0.9687, "step": 457 }, { "epoch": 0.34875309347039785, "grad_norm": 45.766780853271484, "learning_rate": 9.397521377835135e-06, "loss": 0.9491, "step": 458 }, { "epoch": 0.34951456310679613, "grad_norm": 34.62493896484375, "learning_rate": 9.394618470937836e-06, "loss": 0.7288, "step": 459 }, { "epoch": 0.35027603274319435, "grad_norm": 49.47513961791992, "learning_rate": 9.391709037920726e-06, "loss": 0.8846, "step": 460 }, { "epoch": 0.35103750237959264, "grad_norm": 40.401512145996094, "learning_rate": 9.388793083104393e-06, "loss": 0.8286, "step": 461 }, { "epoch": 0.35179897201599086, "grad_norm": 44.801109313964844, "learning_rate": 9.3858706108191e-06, "loss": 0.8382, "step": 462 }, { "epoch": 0.3525604416523891, "grad_norm": 36.950382232666016, "learning_rate": 9.382941625404793e-06, "loss": 0.7518, "step": 463 }, { "epoch": 0.35332191128878737, "grad_norm": 46.91257858276367, "learning_rate": 9.380006131211092e-06, "loss": 0.9471, "step": 464 }, { "epoch": 0.3540833809251856, "grad_norm": 43.36894989013672, "learning_rate": 9.377064132597277e-06, "loss": 0.8558, "step": 465 }, { "epoch": 0.3548448505615839, "grad_norm": 53.6937370300293, "learning_rate": 9.374115633932292e-06, "loss": 0.9714, "step": 466 }, { "epoch": 0.3556063201979821, "grad_norm": 49.51094055175781, "learning_rate": 9.371160639594733e-06, "loss": 0.9696, "step": 467 }, { "epoch": 0.35636778983438033, "grad_norm": 70.25513458251953, "learning_rate": 9.36819915397284e-06, "loss": 1.4756, "step": 468 }, { "epoch": 0.3571292594707786, "grad_norm": 41.462982177734375, "learning_rate": 9.365231181464494e-06, "loss": 0.7885, "step": 469 }, { "epoch": 0.35789072910717684, "grad_norm": 52.57938766479492, "learning_rate": 9.36225672647721e-06, "loss": 1.0787, "step": 470 }, { "epoch": 0.3586521987435751, "grad_norm": 31.510698318481445, "learning_rate": 9.35927579342813e-06, "loss": 0.8847, "step": 471 }, { "epoch": 0.35941366837997335, "grad_norm": 44.495147705078125, "learning_rate": 9.356288386744015e-06, "loss": 0.8841, "step": 472 }, { "epoch": 0.3601751380163716, "grad_norm": 35.463748931884766, "learning_rate": 9.35329451086124e-06, "loss": 0.9095, "step": 473 }, { "epoch": 0.36093660765276986, "grad_norm": 40.00856399536133, "learning_rate": 9.350294170225782e-06, "loss": 0.7667, "step": 474 }, { "epoch": 0.3616980772891681, "grad_norm": 46.01627731323242, "learning_rate": 9.34728736929323e-06, "loss": 0.922, "step": 475 }, { "epoch": 0.36245954692556637, "grad_norm": 39.330169677734375, "learning_rate": 9.344274112528754e-06, "loss": 0.7242, "step": 476 }, { "epoch": 0.3632210165619646, "grad_norm": 41.52873611450195, "learning_rate": 9.34125440440712e-06, "loss": 0.7228, "step": 477 }, { "epoch": 0.3639824861983628, "grad_norm": 33.01096725463867, "learning_rate": 9.338228249412668e-06, "loss": 0.8528, "step": 478 }, { "epoch": 0.3647439558347611, "grad_norm": 36.77853012084961, "learning_rate": 9.33519565203932e-06, "loss": 0.869, "step": 479 }, { "epoch": 0.36550542547115933, "grad_norm": 44.876426696777344, "learning_rate": 9.332156616790552e-06, "loss": 1.0064, "step": 480 }, { "epoch": 0.3662668951075576, "grad_norm": 45.61384963989258, "learning_rate": 9.329111148179417e-06, "loss": 0.8066, "step": 481 }, { "epoch": 0.36702836474395584, "grad_norm": 42.6773796081543, "learning_rate": 9.326059250728508e-06, "loss": 0.839, "step": 482 }, { "epoch": 0.36778983438035406, "grad_norm": 42.04776382446289, "learning_rate": 9.32300092896997e-06, "loss": 0.7902, "step": 483 }, { "epoch": 0.36855130401675235, "grad_norm": 44.73598861694336, "learning_rate": 9.319936187445491e-06, "loss": 0.7914, "step": 484 }, { "epoch": 0.3693127736531506, "grad_norm": 28.97296714782715, "learning_rate": 9.316865030706288e-06, "loss": 0.7196, "step": 485 }, { "epoch": 0.37007424328954885, "grad_norm": 39.61836624145508, "learning_rate": 9.313787463313107e-06, "loss": 0.7217, "step": 486 }, { "epoch": 0.3708357129259471, "grad_norm": 40.51829147338867, "learning_rate": 9.310703489836211e-06, "loss": 0.8674, "step": 487 }, { "epoch": 0.3715971825623453, "grad_norm": 29.400951385498047, "learning_rate": 9.307613114855383e-06, "loss": 0.7207, "step": 488 }, { "epoch": 0.3723586521987436, "grad_norm": 29.25424575805664, "learning_rate": 9.304516342959908e-06, "loss": 0.6184, "step": 489 }, { "epoch": 0.3731201218351418, "grad_norm": 48.31173324584961, "learning_rate": 9.301413178748568e-06, "loss": 0.6037, "step": 490 }, { "epoch": 0.3738815914715401, "grad_norm": 40.35416793823242, "learning_rate": 9.298303626829641e-06, "loss": 0.9935, "step": 491 }, { "epoch": 0.3746430611079383, "grad_norm": 33.47250747680664, "learning_rate": 9.29518769182089e-06, "loss": 0.6797, "step": 492 }, { "epoch": 0.37540453074433655, "grad_norm": 40.839599609375, "learning_rate": 9.292065378349561e-06, "loss": 0.8908, "step": 493 }, { "epoch": 0.37616600038073483, "grad_norm": 99.43079376220703, "learning_rate": 9.288936691052366e-06, "loss": 0.9635, "step": 494 }, { "epoch": 0.37692747001713306, "grad_norm": 37.33384704589844, "learning_rate": 9.285801634575484e-06, "loss": 0.9248, "step": 495 }, { "epoch": 0.37768893965353134, "grad_norm": 43.419593811035156, "learning_rate": 9.282660213574554e-06, "loss": 0.9641, "step": 496 }, { "epoch": 0.37845040928992957, "grad_norm": 37.27656555175781, "learning_rate": 9.279512432714668e-06, "loss": 0.826, "step": 497 }, { "epoch": 0.3792118789263278, "grad_norm": 33.71250534057617, "learning_rate": 9.276358296670357e-06, "loss": 0.6497, "step": 498 }, { "epoch": 0.3799733485627261, "grad_norm": 35.924400329589844, "learning_rate": 9.273197810125595e-06, "loss": 0.6516, "step": 499 }, { "epoch": 0.3807348181991243, "grad_norm": 46.43340301513672, "learning_rate": 9.270030977773785e-06, "loss": 0.7873, "step": 500 }, { "epoch": 0.3814962878355226, "grad_norm": 35.527164459228516, "learning_rate": 9.26685780431775e-06, "loss": 0.8364, "step": 501 }, { "epoch": 0.3822577574719208, "grad_norm": 38.08017349243164, "learning_rate": 9.263678294469738e-06, "loss": 0.8295, "step": 502 }, { "epoch": 0.38301922710831904, "grad_norm": 44.70781326293945, "learning_rate": 9.260492452951399e-06, "loss": 0.7772, "step": 503 }, { "epoch": 0.3837806967447173, "grad_norm": 40.65570831298828, "learning_rate": 9.257300284493788e-06, "loss": 0.8622, "step": 504 }, { "epoch": 0.38454216638111555, "grad_norm": 46.29487228393555, "learning_rate": 9.254101793837357e-06, "loss": 0.9253, "step": 505 }, { "epoch": 0.38530363601751383, "grad_norm": 43.05247497558594, "learning_rate": 9.250896985731943e-06, "loss": 0.7856, "step": 506 }, { "epoch": 0.38606510565391206, "grad_norm": 45.552330017089844, "learning_rate": 9.247685864936773e-06, "loss": 0.9366, "step": 507 }, { "epoch": 0.3868265752903103, "grad_norm": 40.42953872680664, "learning_rate": 9.244468436220437e-06, "loss": 0.8822, "step": 508 }, { "epoch": 0.38758804492670856, "grad_norm": 41.48854446411133, "learning_rate": 9.241244704360902e-06, "loss": 0.9342, "step": 509 }, { "epoch": 0.3883495145631068, "grad_norm": 47.127227783203125, "learning_rate": 9.23801467414549e-06, "loss": 0.8699, "step": 510 }, { "epoch": 0.38911098419950507, "grad_norm": 38.96549606323242, "learning_rate": 9.234778350370877e-06, "loss": 0.8706, "step": 511 }, { "epoch": 0.3898724538359033, "grad_norm": 36.132545471191406, "learning_rate": 9.231535737843087e-06, "loss": 0.6892, "step": 512 }, { "epoch": 0.3906339234723015, "grad_norm": 42.765647888183594, "learning_rate": 9.22828684137748e-06, "loss": 0.8688, "step": 513 }, { "epoch": 0.3913953931086998, "grad_norm": 48.11317825317383, "learning_rate": 9.225031665798754e-06, "loss": 0.9007, "step": 514 }, { "epoch": 0.39215686274509803, "grad_norm": 40.7263298034668, "learning_rate": 9.221770215940921e-06, "loss": 0.8925, "step": 515 }, { "epoch": 0.39291833238149626, "grad_norm": 38.51908493041992, "learning_rate": 9.218502496647323e-06, "loss": 0.8315, "step": 516 }, { "epoch": 0.39367980201789454, "grad_norm": 42.210105895996094, "learning_rate": 9.215228512770601e-06, "loss": 0.8786, "step": 517 }, { "epoch": 0.39444127165429277, "grad_norm": 38.10415267944336, "learning_rate": 9.211948269172705e-06, "loss": 0.8275, "step": 518 }, { "epoch": 0.39520274129069105, "grad_norm": 40.787353515625, "learning_rate": 9.208661770724884e-06, "loss": 0.7724, "step": 519 }, { "epoch": 0.3959642109270893, "grad_norm": 33.491241455078125, "learning_rate": 9.205369022307667e-06, "loss": 0.7493, "step": 520 }, { "epoch": 0.3967256805634875, "grad_norm": 34.938289642333984, "learning_rate": 9.202070028810868e-06, "loss": 0.6538, "step": 521 }, { "epoch": 0.3974871501998858, "grad_norm": 34.13314437866211, "learning_rate": 9.198764795133579e-06, "loss": 0.6973, "step": 522 }, { "epoch": 0.398248619836284, "grad_norm": 47.31309127807617, "learning_rate": 9.195453326184157e-06, "loss": 0.9191, "step": 523 }, { "epoch": 0.3990100894726823, "grad_norm": 31.99027442932129, "learning_rate": 9.192135626880213e-06, "loss": 0.6122, "step": 524 }, { "epoch": 0.3997715591090805, "grad_norm": 35.60274124145508, "learning_rate": 9.188811702148618e-06, "loss": 0.7511, "step": 525 }, { "epoch": 0.40053302874547875, "grad_norm": 38.91366195678711, "learning_rate": 9.185481556925483e-06, "loss": 0.6532, "step": 526 }, { "epoch": 0.40129449838187703, "grad_norm": 39.65351104736328, "learning_rate": 9.182145196156159e-06, "loss": 0.8747, "step": 527 }, { "epoch": 0.40205596801827526, "grad_norm": 39.787052154541016, "learning_rate": 9.178802624795222e-06, "loss": 0.9385, "step": 528 }, { "epoch": 0.40281743765467354, "grad_norm": 40.41245651245117, "learning_rate": 9.175453847806482e-06, "loss": 0.8325, "step": 529 }, { "epoch": 0.40357890729107176, "grad_norm": 33.6597900390625, "learning_rate": 9.172098870162953e-06, "loss": 0.7872, "step": 530 }, { "epoch": 0.40434037692747, "grad_norm": 45.033607482910156, "learning_rate": 9.168737696846866e-06, "loss": 1.0956, "step": 531 }, { "epoch": 0.4051018465638683, "grad_norm": 50.52238464355469, "learning_rate": 9.165370332849642e-06, "loss": 0.5881, "step": 532 }, { "epoch": 0.4058633162002665, "grad_norm": 35.1140251159668, "learning_rate": 9.161996783171908e-06, "loss": 0.6781, "step": 533 }, { "epoch": 0.4066247858366648, "grad_norm": 32.561767578125, "learning_rate": 9.158617052823469e-06, "loss": 0.6794, "step": 534 }, { "epoch": 0.407386255473063, "grad_norm": 32.99089050292969, "learning_rate": 9.155231146823309e-06, "loss": 0.6522, "step": 535 }, { "epoch": 0.40814772510946123, "grad_norm": 32.225341796875, "learning_rate": 9.151839070199585e-06, "loss": 0.875, "step": 536 }, { "epoch": 0.4089091947458595, "grad_norm": 31.30025291442871, "learning_rate": 9.148440827989615e-06, "loss": 0.6423, "step": 537 }, { "epoch": 0.40967066438225774, "grad_norm": 32.41795349121094, "learning_rate": 9.145036425239877e-06, "loss": 0.6811, "step": 538 }, { "epoch": 0.410432134018656, "grad_norm": 48.22577667236328, "learning_rate": 9.141625867005997e-06, "loss": 0.7582, "step": 539 }, { "epoch": 0.41119360365505425, "grad_norm": 42.18339920043945, "learning_rate": 9.138209158352736e-06, "loss": 0.8656, "step": 540 }, { "epoch": 0.4119550732914525, "grad_norm": 36.78019332885742, "learning_rate": 9.134786304353998e-06, "loss": 0.9369, "step": 541 }, { "epoch": 0.41271654292785076, "grad_norm": 29.20226287841797, "learning_rate": 9.131357310092803e-06, "loss": 0.7197, "step": 542 }, { "epoch": 0.413478012564249, "grad_norm": 35.52668762207031, "learning_rate": 9.127922180661297e-06, "loss": 0.7683, "step": 543 }, { "epoch": 0.41423948220064727, "grad_norm": 38.33584976196289, "learning_rate": 9.124480921160735e-06, "loss": 0.8918, "step": 544 }, { "epoch": 0.4150009518370455, "grad_norm": 33.05094528198242, "learning_rate": 9.121033536701471e-06, "loss": 0.7041, "step": 545 }, { "epoch": 0.4157624214734437, "grad_norm": 58.042510986328125, "learning_rate": 9.117580032402962e-06, "loss": 0.6833, "step": 546 }, { "epoch": 0.416523891109842, "grad_norm": 32.62454605102539, "learning_rate": 9.114120413393748e-06, "loss": 0.7837, "step": 547 }, { "epoch": 0.41728536074624023, "grad_norm": 37.49154281616211, "learning_rate": 9.11065468481145e-06, "loss": 0.7867, "step": 548 }, { "epoch": 0.4180468303826385, "grad_norm": 40.316165924072266, "learning_rate": 9.107182851802765e-06, "loss": 1.0259, "step": 549 }, { "epoch": 0.41880830001903674, "grad_norm": 40.33757781982422, "learning_rate": 9.103704919523453e-06, "loss": 0.7735, "step": 550 }, { "epoch": 0.41956976965543497, "grad_norm": 64.47599792480469, "learning_rate": 9.100220893138327e-06, "loss": 0.8922, "step": 551 }, { "epoch": 0.42033123929183325, "grad_norm": 67.62235260009766, "learning_rate": 9.096730777821261e-06, "loss": 0.941, "step": 552 }, { "epoch": 0.4210927089282315, "grad_norm": 32.08481979370117, "learning_rate": 9.09323457875516e-06, "loss": 0.7371, "step": 553 }, { "epoch": 0.42185417856462976, "grad_norm": 42.49589538574219, "learning_rate": 9.089732301131971e-06, "loss": 0.8789, "step": 554 }, { "epoch": 0.422615648201028, "grad_norm": 42.37563705444336, "learning_rate": 9.086223950152663e-06, "loss": 0.8759, "step": 555 }, { "epoch": 0.4233771178374262, "grad_norm": 33.896888732910156, "learning_rate": 9.082709531027228e-06, "loss": 0.9402, "step": 556 }, { "epoch": 0.4241385874738245, "grad_norm": 40.62684631347656, "learning_rate": 9.079189048974663e-06, "loss": 0.6872, "step": 557 }, { "epoch": 0.4249000571102227, "grad_norm": 40.33720016479492, "learning_rate": 9.07566250922298e-06, "loss": 0.879, "step": 558 }, { "epoch": 0.425661526746621, "grad_norm": 33.24203109741211, "learning_rate": 9.072129917009175e-06, "loss": 0.6897, "step": 559 }, { "epoch": 0.4264229963830192, "grad_norm": 47.51837921142578, "learning_rate": 9.068591277579238e-06, "loss": 0.723, "step": 560 }, { "epoch": 0.42718446601941745, "grad_norm": 35.74695587158203, "learning_rate": 9.06504659618814e-06, "loss": 0.7811, "step": 561 }, { "epoch": 0.42794593565581573, "grad_norm": 32.98100280761719, "learning_rate": 9.06149587809982e-06, "loss": 0.9302, "step": 562 }, { "epoch": 0.42870740529221396, "grad_norm": 30.54879379272461, "learning_rate": 9.057939128587186e-06, "loss": 0.6805, "step": 563 }, { "epoch": 0.42946887492861224, "grad_norm": 58.388309478759766, "learning_rate": 9.054376352932102e-06, "loss": 1.2049, "step": 564 }, { "epoch": 0.43023034456501047, "grad_norm": 39.34674072265625, "learning_rate": 9.050807556425376e-06, "loss": 0.7732, "step": 565 }, { "epoch": 0.4309918142014087, "grad_norm": 68.59402465820312, "learning_rate": 9.047232744366767e-06, "loss": 0.9679, "step": 566 }, { "epoch": 0.431753283837807, "grad_norm": 38.54568099975586, "learning_rate": 9.043651922064957e-06, "loss": 0.7851, "step": 567 }, { "epoch": 0.4325147534742052, "grad_norm": 42.376461029052734, "learning_rate": 9.04006509483756e-06, "loss": 0.8066, "step": 568 }, { "epoch": 0.4332762231106035, "grad_norm": 27.669086456298828, "learning_rate": 9.036472268011107e-06, "loss": 0.5173, "step": 569 }, { "epoch": 0.4340376927470017, "grad_norm": 39.31504821777344, "learning_rate": 9.032873446921031e-06, "loss": 0.8415, "step": 570 }, { "epoch": 0.43479916238339994, "grad_norm": 32.41474533081055, "learning_rate": 9.02926863691168e-06, "loss": 0.7695, "step": 571 }, { "epoch": 0.4355606320197982, "grad_norm": 41.595359802246094, "learning_rate": 9.025657843336283e-06, "loss": 0.8801, "step": 572 }, { "epoch": 0.43632210165619645, "grad_norm": 42.80996322631836, "learning_rate": 9.022041071556962e-06, "loss": 0.9462, "step": 573 }, { "epoch": 0.43708357129259473, "grad_norm": 45.28131103515625, "learning_rate": 9.018418326944713e-06, "loss": 0.6996, "step": 574 }, { "epoch": 0.43784504092899296, "grad_norm": 43.1144905090332, "learning_rate": 9.014789614879406e-06, "loss": 0.9213, "step": 575 }, { "epoch": 0.4386065105653912, "grad_norm": 41.01248550415039, "learning_rate": 9.011154940749769e-06, "loss": 0.788, "step": 576 }, { "epoch": 0.43936798020178947, "grad_norm": 30.024755477905273, "learning_rate": 9.007514309953387e-06, "loss": 0.8227, "step": 577 }, { "epoch": 0.4401294498381877, "grad_norm": 36.76607894897461, "learning_rate": 9.003867727896684e-06, "loss": 0.7102, "step": 578 }, { "epoch": 0.440890919474586, "grad_norm": 39.9501953125, "learning_rate": 9.000215199994933e-06, "loss": 0.9365, "step": 579 }, { "epoch": 0.4416523891109842, "grad_norm": 49.102298736572266, "learning_rate": 8.996556731672228e-06, "loss": 1.0171, "step": 580 }, { "epoch": 0.4424138587473824, "grad_norm": 39.701072692871094, "learning_rate": 8.992892328361485e-06, "loss": 0.7041, "step": 581 }, { "epoch": 0.4431753283837807, "grad_norm": 42.883201599121094, "learning_rate": 8.989221995504437e-06, "loss": 0.8881, "step": 582 }, { "epoch": 0.44393679802017894, "grad_norm": 35.93183135986328, "learning_rate": 8.985545738551625e-06, "loss": 0.849, "step": 583 }, { "epoch": 0.4446982676565772, "grad_norm": 33.20316696166992, "learning_rate": 8.98186356296238e-06, "loss": 0.8081, "step": 584 }, { "epoch": 0.44545973729297544, "grad_norm": 34.129783630371094, "learning_rate": 8.978175474204827e-06, "loss": 0.6737, "step": 585 }, { "epoch": 0.44622120692937367, "grad_norm": 35.47514343261719, "learning_rate": 8.97448147775587e-06, "loss": 0.8261, "step": 586 }, { "epoch": 0.44698267656577195, "grad_norm": 34.96803665161133, "learning_rate": 8.970781579101188e-06, "loss": 0.877, "step": 587 }, { "epoch": 0.4477441462021702, "grad_norm": 36.19900894165039, "learning_rate": 8.967075783735226e-06, "loss": 0.5846, "step": 588 }, { "epoch": 0.44850561583856846, "grad_norm": 35.287315368652344, "learning_rate": 8.963364097161182e-06, "loss": 0.783, "step": 589 }, { "epoch": 0.4492670854749667, "grad_norm": 37.13393783569336, "learning_rate": 8.959646524891005e-06, "loss": 0.6996, "step": 590 }, { "epoch": 0.4500285551113649, "grad_norm": 35.42992401123047, "learning_rate": 8.955923072445386e-06, "loss": 0.7812, "step": 591 }, { "epoch": 0.4507900247477632, "grad_norm": 34.43177032470703, "learning_rate": 8.952193745353743e-06, "loss": 0.6311, "step": 592 }, { "epoch": 0.4515514943841614, "grad_norm": 29.86054229736328, "learning_rate": 8.948458549154224e-06, "loss": 0.5255, "step": 593 }, { "epoch": 0.4523129640205597, "grad_norm": 40.26968765258789, "learning_rate": 8.94471748939369e-06, "loss": 0.8848, "step": 594 }, { "epoch": 0.45307443365695793, "grad_norm": 35.20211410522461, "learning_rate": 8.940970571627707e-06, "loss": 0.6966, "step": 595 }, { "epoch": 0.45383590329335616, "grad_norm": 27.7930850982666, "learning_rate": 8.937217801420544e-06, "loss": 0.7314, "step": 596 }, { "epoch": 0.45459737292975444, "grad_norm": 34.17549514770508, "learning_rate": 8.933459184345162e-06, "loss": 0.8493, "step": 597 }, { "epoch": 0.45535884256615267, "grad_norm": 32.22350311279297, "learning_rate": 8.9296947259832e-06, "loss": 0.8395, "step": 598 }, { "epoch": 0.45612031220255095, "grad_norm": 26.227863311767578, "learning_rate": 8.925924431924974e-06, "loss": 0.4723, "step": 599 }, { "epoch": 0.4568817818389492, "grad_norm": 32.933223724365234, "learning_rate": 8.922148307769469e-06, "loss": 0.7971, "step": 600 }, { "epoch": 0.4576432514753474, "grad_norm": 43.23099899291992, "learning_rate": 8.918366359124322e-06, "loss": 0.8534, "step": 601 }, { "epoch": 0.4584047211117457, "grad_norm": 49.473995208740234, "learning_rate": 8.914578591605824e-06, "loss": 0.9335, "step": 602 }, { "epoch": 0.4591661907481439, "grad_norm": 31.68277359008789, "learning_rate": 8.910785010838905e-06, "loss": 0.8195, "step": 603 }, { "epoch": 0.4599276603845422, "grad_norm": 38.720611572265625, "learning_rate": 8.906985622457129e-06, "loss": 1.0383, "step": 604 }, { "epoch": 0.4606891300209404, "grad_norm": 32.30598831176758, "learning_rate": 8.903180432102686e-06, "loss": 0.8484, "step": 605 }, { "epoch": 0.46145059965733864, "grad_norm": 35.373565673828125, "learning_rate": 8.899369445426377e-06, "loss": 0.7706, "step": 606 }, { "epoch": 0.4622120692937369, "grad_norm": 36.211456298828125, "learning_rate": 8.895552668087615e-06, "loss": 0.7762, "step": 607 }, { "epoch": 0.46297353893013515, "grad_norm": 31.659029006958008, "learning_rate": 8.891730105754412e-06, "loss": 0.6671, "step": 608 }, { "epoch": 0.46373500856653344, "grad_norm": 34.945106506347656, "learning_rate": 8.887901764103368e-06, "loss": 0.9202, "step": 609 }, { "epoch": 0.46449647820293166, "grad_norm": 31.101940155029297, "learning_rate": 8.88406764881967e-06, "loss": 0.7485, "step": 610 }, { "epoch": 0.4652579478393299, "grad_norm": 39.410953521728516, "learning_rate": 8.880227765597074e-06, "loss": 1.0098, "step": 611 }, { "epoch": 0.46601941747572817, "grad_norm": 30.449037551879883, "learning_rate": 8.876382120137906e-06, "loss": 0.7136, "step": 612 }, { "epoch": 0.4667808871121264, "grad_norm": 42.712135314941406, "learning_rate": 8.872530718153044e-06, "loss": 0.9093, "step": 613 }, { "epoch": 0.4675423567485247, "grad_norm": 30.30866050720215, "learning_rate": 8.86867356536192e-06, "loss": 0.6193, "step": 614 }, { "epoch": 0.4683038263849229, "grad_norm": 38.80727005004883, "learning_rate": 8.864810667492502e-06, "loss": 0.496, "step": 615 }, { "epoch": 0.46906529602132113, "grad_norm": 37.91781234741211, "learning_rate": 8.860942030281294e-06, "loss": 0.8192, "step": 616 }, { "epoch": 0.4698267656577194, "grad_norm": 35.720909118652344, "learning_rate": 8.857067659473319e-06, "loss": 0.897, "step": 617 }, { "epoch": 0.47058823529411764, "grad_norm": 40.22643280029297, "learning_rate": 8.853187560822113e-06, "loss": 0.7797, "step": 618 }, { "epoch": 0.4713497049305159, "grad_norm": 36.7399787902832, "learning_rate": 8.849301740089725e-06, "loss": 0.8, "step": 619 }, { "epoch": 0.47211117456691415, "grad_norm": 34.08172607421875, "learning_rate": 8.845410203046695e-06, "loss": 0.8473, "step": 620 }, { "epoch": 0.4728726442033124, "grad_norm": 45.90078353881836, "learning_rate": 8.841512955472054e-06, "loss": 1.0885, "step": 621 }, { "epoch": 0.47363411383971066, "grad_norm": 40.18413543701172, "learning_rate": 8.837610003153316e-06, "loss": 0.7083, "step": 622 }, { "epoch": 0.4743955834761089, "grad_norm": 43.39189910888672, "learning_rate": 8.83370135188646e-06, "loss": 0.963, "step": 623 }, { "epoch": 0.47515705311250717, "grad_norm": 27.63300895690918, "learning_rate": 8.829787007475936e-06, "loss": 0.774, "step": 624 }, { "epoch": 0.4759185227489054, "grad_norm": 29.548891067504883, "learning_rate": 8.825866975734642e-06, "loss": 0.5855, "step": 625 }, { "epoch": 0.4766799923853036, "grad_norm": 55.26377487182617, "learning_rate": 8.821941262483924e-06, "loss": 1.2597, "step": 626 }, { "epoch": 0.4774414620217019, "grad_norm": 36.47280502319336, "learning_rate": 8.81800987355357e-06, "loss": 0.8026, "step": 627 }, { "epoch": 0.4782029316581001, "grad_norm": 31.497495651245117, "learning_rate": 8.814072814781788e-06, "loss": 0.7601, "step": 628 }, { "epoch": 0.47896440129449835, "grad_norm": 37.276546478271484, "learning_rate": 8.810130092015211e-06, "loss": 0.8585, "step": 629 }, { "epoch": 0.47972587093089664, "grad_norm": 31.686859130859375, "learning_rate": 8.806181711108883e-06, "loss": 0.8257, "step": 630 }, { "epoch": 0.48048734056729486, "grad_norm": 45.68613815307617, "learning_rate": 8.802227677926249e-06, "loss": 0.774, "step": 631 }, { "epoch": 0.48124881020369314, "grad_norm": 30.245437622070312, "learning_rate": 8.798267998339149e-06, "loss": 0.75, "step": 632 }, { "epoch": 0.48201027984009137, "grad_norm": 26.60959815979004, "learning_rate": 8.794302678227809e-06, "loss": 0.592, "step": 633 }, { "epoch": 0.4827717494764896, "grad_norm": 32.366424560546875, "learning_rate": 8.790331723480828e-06, "loss": 0.6491, "step": 634 }, { "epoch": 0.4835332191128879, "grad_norm": 35.88400650024414, "learning_rate": 8.786355139995173e-06, "loss": 0.7825, "step": 635 }, { "epoch": 0.4842946887492861, "grad_norm": 33.99674606323242, "learning_rate": 8.782372933676174e-06, "loss": 0.7043, "step": 636 }, { "epoch": 0.4850561583856844, "grad_norm": 30.469955444335938, "learning_rate": 8.778385110437507e-06, "loss": 0.8757, "step": 637 }, { "epoch": 0.4858176280220826, "grad_norm": 32.127506256103516, "learning_rate": 8.77439167620119e-06, "loss": 0.7006, "step": 638 }, { "epoch": 0.48657909765848084, "grad_norm": 33.425872802734375, "learning_rate": 8.770392636897576e-06, "loss": 0.8351, "step": 639 }, { "epoch": 0.4873405672948791, "grad_norm": 38.01253128051758, "learning_rate": 8.766387998465337e-06, "loss": 0.962, "step": 640 }, { "epoch": 0.48810203693127735, "grad_norm": 42.143707275390625, "learning_rate": 8.762377766851467e-06, "loss": 0.9168, "step": 641 }, { "epoch": 0.48886350656767563, "grad_norm": 32.78673553466797, "learning_rate": 8.758361948011256e-06, "loss": 0.7231, "step": 642 }, { "epoch": 0.48962497620407386, "grad_norm": 41.47932815551758, "learning_rate": 8.754340547908301e-06, "loss": 0.842, "step": 643 }, { "epoch": 0.4903864458404721, "grad_norm": 30.17685317993164, "learning_rate": 8.750313572514482e-06, "loss": 0.7623, "step": 644 }, { "epoch": 0.49114791547687037, "grad_norm": 44.7989616394043, "learning_rate": 8.746281027809958e-06, "loss": 0.9316, "step": 645 }, { "epoch": 0.4919093851132686, "grad_norm": 33.90018844604492, "learning_rate": 8.742242919783161e-06, "loss": 0.8259, "step": 646 }, { "epoch": 0.4926708547496669, "grad_norm": 28.30022621154785, "learning_rate": 8.738199254430785e-06, "loss": 0.747, "step": 647 }, { "epoch": 0.4934323243860651, "grad_norm": 24.467458724975586, "learning_rate": 8.734150037757774e-06, "loss": 0.6057, "step": 648 }, { "epoch": 0.49419379402246333, "grad_norm": 38.404075622558594, "learning_rate": 8.730095275777317e-06, "loss": 0.8806, "step": 649 }, { "epoch": 0.4949552636588616, "grad_norm": 37.866458892822266, "learning_rate": 8.72603497451084e-06, "loss": 0.7884, "step": 650 }, { "epoch": 0.49571673329525984, "grad_norm": 34.43553924560547, "learning_rate": 8.721969139987994e-06, "loss": 0.7668, "step": 651 }, { "epoch": 0.4964782029316581, "grad_norm": 30.790082931518555, "learning_rate": 8.717897778246642e-06, "loss": 0.8646, "step": 652 }, { "epoch": 0.49723967256805635, "grad_norm": 32.69529724121094, "learning_rate": 8.713820895332864e-06, "loss": 0.8441, "step": 653 }, { "epoch": 0.49800114220445457, "grad_norm": 31.344131469726562, "learning_rate": 8.709738497300932e-06, "loss": 0.7419, "step": 654 }, { "epoch": 0.49876261184085285, "grad_norm": 34.3707275390625, "learning_rate": 8.70565059021331e-06, "loss": 0.9512, "step": 655 }, { "epoch": 0.4995240814772511, "grad_norm": 30.182653427124023, "learning_rate": 8.701557180140647e-06, "loss": 0.8681, "step": 656 }, { "epoch": 0.5002855511136494, "grad_norm": 35.306846618652344, "learning_rate": 8.697458273161757e-06, "loss": 1.0412, "step": 657 }, { "epoch": 0.5010470207500476, "grad_norm": 31.0938720703125, "learning_rate": 8.693353875363623e-06, "loss": 0.692, "step": 658 }, { "epoch": 0.5018084903864458, "grad_norm": 31.533510208129883, "learning_rate": 8.689243992841379e-06, "loss": 0.6184, "step": 659 }, { "epoch": 0.502569960022844, "grad_norm": 33.87437438964844, "learning_rate": 8.685128631698304e-06, "loss": 0.7929, "step": 660 }, { "epoch": 0.5033314296592424, "grad_norm": 33.65624237060547, "learning_rate": 8.681007798045814e-06, "loss": 0.9535, "step": 661 }, { "epoch": 0.5040928992956406, "grad_norm": 33.318153381347656, "learning_rate": 8.676881498003453e-06, "loss": 0.8338, "step": 662 }, { "epoch": 0.5048543689320388, "grad_norm": 32.10042953491211, "learning_rate": 8.672749737698878e-06, "loss": 0.7802, "step": 663 }, { "epoch": 0.5056158385684371, "grad_norm": 43.71488571166992, "learning_rate": 8.668612523267862e-06, "loss": 0.9463, "step": 664 }, { "epoch": 0.5063773082048353, "grad_norm": 41.089439392089844, "learning_rate": 8.66446986085427e-06, "loss": 1.0539, "step": 665 }, { "epoch": 0.5071387778412336, "grad_norm": 28.69977378845215, "learning_rate": 8.66032175661006e-06, "loss": 0.7834, "step": 666 }, { "epoch": 0.5079002474776318, "grad_norm": 37.99454116821289, "learning_rate": 8.656168216695276e-06, "loss": 0.8248, "step": 667 }, { "epoch": 0.5086617171140301, "grad_norm": 35.04216384887695, "learning_rate": 8.652009247278026e-06, "loss": 0.7818, "step": 668 }, { "epoch": 0.5094231867504283, "grad_norm": 29.913171768188477, "learning_rate": 8.647844854534487e-06, "loss": 0.6774, "step": 669 }, { "epoch": 0.5101846563868265, "grad_norm": 31.227197647094727, "learning_rate": 8.643675044648885e-06, "loss": 0.7429, "step": 670 }, { "epoch": 0.5109461260232249, "grad_norm": 33.46771240234375, "learning_rate": 8.639499823813499e-06, "loss": 0.8269, "step": 671 }, { "epoch": 0.5117075956596231, "grad_norm": 30.12987518310547, "learning_rate": 8.635319198228633e-06, "loss": 0.6747, "step": 672 }, { "epoch": 0.5124690652960213, "grad_norm": 33.5628662109375, "learning_rate": 8.631133174102623e-06, "loss": 0.7735, "step": 673 }, { "epoch": 0.5132305349324195, "grad_norm": 30.92879295349121, "learning_rate": 8.62694175765182e-06, "loss": 0.7511, "step": 674 }, { "epoch": 0.5139920045688178, "grad_norm": 29.685874938964844, "learning_rate": 8.622744955100589e-06, "loss": 0.7242, "step": 675 }, { "epoch": 0.5147534742052161, "grad_norm": 27.040605545043945, "learning_rate": 8.618542772681282e-06, "loss": 0.6924, "step": 676 }, { "epoch": 0.5155149438416143, "grad_norm": 36.78512954711914, "learning_rate": 8.614335216634247e-06, "loss": 0.7219, "step": 677 }, { "epoch": 0.5162764134780126, "grad_norm": 39.6012077331543, "learning_rate": 8.610122293207815e-06, "loss": 0.8196, "step": 678 }, { "epoch": 0.5170378831144108, "grad_norm": 33.704833984375, "learning_rate": 8.60590400865828e-06, "loss": 0.9052, "step": 679 }, { "epoch": 0.517799352750809, "grad_norm": 35.91033172607422, "learning_rate": 8.601680369249905e-06, "loss": 0.8486, "step": 680 }, { "epoch": 0.5185608223872074, "grad_norm": 34.80812072753906, "learning_rate": 8.5974513812549e-06, "loss": 0.7627, "step": 681 }, { "epoch": 0.5193222920236056, "grad_norm": 34.02023696899414, "learning_rate": 8.593217050953417e-06, "loss": 0.6358, "step": 682 }, { "epoch": 0.5200837616600038, "grad_norm": 40.504356384277344, "learning_rate": 8.588977384633549e-06, "loss": 0.8517, "step": 683 }, { "epoch": 0.520845231296402, "grad_norm": 38.43982696533203, "learning_rate": 8.584732388591303e-06, "loss": 0.8482, "step": 684 }, { "epoch": 0.5216067009328003, "grad_norm": 31.85175323486328, "learning_rate": 8.580482069130609e-06, "loss": 0.8291, "step": 685 }, { "epoch": 0.5223681705691986, "grad_norm": 43.681121826171875, "learning_rate": 8.576226432563298e-06, "loss": 0.9371, "step": 686 }, { "epoch": 0.5231296402055968, "grad_norm": 37.21615982055664, "learning_rate": 8.571965485209098e-06, "loss": 0.8825, "step": 687 }, { "epoch": 0.523891109841995, "grad_norm": 34.33724594116211, "learning_rate": 8.56769923339562e-06, "loss": 0.8466, "step": 688 }, { "epoch": 0.5246525794783933, "grad_norm": 28.02695655822754, "learning_rate": 8.563427683458365e-06, "loss": 0.6877, "step": 689 }, { "epoch": 0.5254140491147915, "grad_norm": 38.37015914916992, "learning_rate": 8.559150841740686e-06, "loss": 0.9541, "step": 690 }, { "epoch": 0.5261755187511898, "grad_norm": 28.946571350097656, "learning_rate": 8.554868714593802e-06, "loss": 0.6303, "step": 691 }, { "epoch": 0.5269369883875881, "grad_norm": 34.604644775390625, "learning_rate": 8.550581308376783e-06, "loss": 0.8105, "step": 692 }, { "epoch": 0.5276984580239863, "grad_norm": 33.502750396728516, "learning_rate": 8.546288629456535e-06, "loss": 0.7431, "step": 693 }, { "epoch": 0.5284599276603845, "grad_norm": 33.93081283569336, "learning_rate": 8.541990684207793e-06, "loss": 0.8172, "step": 694 }, { "epoch": 0.5292213972967827, "grad_norm": 42.208980560302734, "learning_rate": 8.537687479013118e-06, "loss": 0.6655, "step": 695 }, { "epoch": 0.5299828669331811, "grad_norm": 31.920230865478516, "learning_rate": 8.533379020262874e-06, "loss": 0.7987, "step": 696 }, { "epoch": 0.5307443365695793, "grad_norm": 36.94475555419922, "learning_rate": 8.529065314355237e-06, "loss": 0.9525, "step": 697 }, { "epoch": 0.5315058062059775, "grad_norm": 31.45237159729004, "learning_rate": 8.524746367696165e-06, "loss": 0.629, "step": 698 }, { "epoch": 0.5322672758423758, "grad_norm": 27.37586784362793, "learning_rate": 8.520422186699405e-06, "loss": 0.7257, "step": 699 }, { "epoch": 0.533028745478774, "grad_norm": 23.65913963317871, "learning_rate": 8.516092777786475e-06, "loss": 0.6115, "step": 700 }, { "epoch": 0.5337902151151723, "grad_norm": 29.19643211364746, "learning_rate": 8.511758147386659e-06, "loss": 0.8932, "step": 701 }, { "epoch": 0.5345516847515706, "grad_norm": 36.52588653564453, "learning_rate": 8.50741830193699e-06, "loss": 0.8328, "step": 702 }, { "epoch": 0.5353131543879688, "grad_norm": 36.54971694946289, "learning_rate": 8.503073247882252e-06, "loss": 0.9755, "step": 703 }, { "epoch": 0.536074624024367, "grad_norm": 42.013641357421875, "learning_rate": 8.498722991674955e-06, "loss": 1.0506, "step": 704 }, { "epoch": 0.5368360936607652, "grad_norm": 36.055870056152344, "learning_rate": 8.494367539775345e-06, "loss": 0.8876, "step": 705 }, { "epoch": 0.5375975632971636, "grad_norm": 30.088356018066406, "learning_rate": 8.490006898651377e-06, "loss": 0.7966, "step": 706 }, { "epoch": 0.5383590329335618, "grad_norm": 40.09549331665039, "learning_rate": 8.485641074778714e-06, "loss": 0.8852, "step": 707 }, { "epoch": 0.53912050256996, "grad_norm": 31.852556228637695, "learning_rate": 8.48127007464071e-06, "loss": 0.7979, "step": 708 }, { "epoch": 0.5398819722063583, "grad_norm": 35.1148567199707, "learning_rate": 8.47689390472842e-06, "loss": 0.7933, "step": 709 }, { "epoch": 0.5406434418427565, "grad_norm": 26.733436584472656, "learning_rate": 8.472512571540558e-06, "loss": 0.4899, "step": 710 }, { "epoch": 0.5414049114791548, "grad_norm": 44.820186614990234, "learning_rate": 8.468126081583524e-06, "loss": 0.7472, "step": 711 }, { "epoch": 0.542166381115553, "grad_norm": 34.0068359375, "learning_rate": 8.463734441371359e-06, "loss": 0.7348, "step": 712 }, { "epoch": 0.5429278507519513, "grad_norm": 37.57758712768555, "learning_rate": 8.459337657425762e-06, "loss": 0.7294, "step": 713 }, { "epoch": 0.5436893203883495, "grad_norm": 37.07135009765625, "learning_rate": 8.454935736276071e-06, "loss": 0.7843, "step": 714 }, { "epoch": 0.5444507900247477, "grad_norm": 37.17467498779297, "learning_rate": 8.450528684459249e-06, "loss": 0.7833, "step": 715 }, { "epoch": 0.5452122596611461, "grad_norm": 27.351919174194336, "learning_rate": 8.44611650851988e-06, "loss": 0.6807, "step": 716 }, { "epoch": 0.5459737292975443, "grad_norm": 34.79734420776367, "learning_rate": 8.441699215010157e-06, "loss": 0.6943, "step": 717 }, { "epoch": 0.5467351989339425, "grad_norm": 34.36366271972656, "learning_rate": 8.437276810489871e-06, "loss": 0.7774, "step": 718 }, { "epoch": 0.5474966685703407, "grad_norm": 31.424224853515625, "learning_rate": 8.432849301526406e-06, "loss": 0.8368, "step": 719 }, { "epoch": 0.548258138206739, "grad_norm": 33.47733688354492, "learning_rate": 8.428416694694728e-06, "loss": 0.7105, "step": 720 }, { "epoch": 0.5490196078431373, "grad_norm": 27.397165298461914, "learning_rate": 8.423978996577367e-06, "loss": 0.674, "step": 721 }, { "epoch": 0.5497810774795355, "grad_norm": 30.114990234375, "learning_rate": 8.419536213764417e-06, "loss": 0.7754, "step": 722 }, { "epoch": 0.5505425471159338, "grad_norm": 29.79185676574707, "learning_rate": 8.415088352853525e-06, "loss": 0.7032, "step": 723 }, { "epoch": 0.551304016752332, "grad_norm": 33.54899597167969, "learning_rate": 8.410635420449877e-06, "loss": 0.7938, "step": 724 }, { "epoch": 0.5520654863887302, "grad_norm": 41.448951721191406, "learning_rate": 8.406177423166188e-06, "loss": 0.763, "step": 725 }, { "epoch": 0.5528269560251285, "grad_norm": 28.263797760009766, "learning_rate": 8.4017143676227e-06, "loss": 0.7605, "step": 726 }, { "epoch": 0.5535884256615268, "grad_norm": 26.69054412841797, "learning_rate": 8.397246260447162e-06, "loss": 0.6578, "step": 727 }, { "epoch": 0.554349895297925, "grad_norm": 24.359060287475586, "learning_rate": 8.392773108274827e-06, "loss": 0.5786, "step": 728 }, { "epoch": 0.5551113649343232, "grad_norm": 32.5286979675293, "learning_rate": 8.388294917748439e-06, "loss": 0.8173, "step": 729 }, { "epoch": 0.5558728345707215, "grad_norm": 27.81627082824707, "learning_rate": 8.383811695518224e-06, "loss": 0.7195, "step": 730 }, { "epoch": 0.5566343042071198, "grad_norm": 27.911325454711914, "learning_rate": 8.379323448241881e-06, "loss": 0.7935, "step": 731 }, { "epoch": 0.557395773843518, "grad_norm": 28.142335891723633, "learning_rate": 8.37483018258457e-06, "loss": 0.8224, "step": 732 }, { "epoch": 0.5581572434799162, "grad_norm": 26.930925369262695, "learning_rate": 8.37033190521891e-06, "loss": 0.7377, "step": 733 }, { "epoch": 0.5589187131163145, "grad_norm": 27.190326690673828, "learning_rate": 8.365828622824948e-06, "loss": 0.7584, "step": 734 }, { "epoch": 0.5596801827527127, "grad_norm": 34.811767578125, "learning_rate": 8.361320342090177e-06, "loss": 0.8059, "step": 735 }, { "epoch": 0.560441652389111, "grad_norm": 27.499380111694336, "learning_rate": 8.35680706970951e-06, "loss": 0.7386, "step": 736 }, { "epoch": 0.5612031220255093, "grad_norm": 27.849960327148438, "learning_rate": 8.352288812385268e-06, "loss": 0.8627, "step": 737 }, { "epoch": 0.5619645916619075, "grad_norm": 38.18281936645508, "learning_rate": 8.34776557682718e-06, "loss": 0.6196, "step": 738 }, { "epoch": 0.5627260612983057, "grad_norm": 42.69196701049805, "learning_rate": 8.343237369752361e-06, "loss": 0.9305, "step": 739 }, { "epoch": 0.5634875309347039, "grad_norm": 34.84543228149414, "learning_rate": 8.338704197885319e-06, "loss": 0.8611, "step": 740 }, { "epoch": 0.5642490005711023, "grad_norm": 31.243303298950195, "learning_rate": 8.334166067957925e-06, "loss": 0.7997, "step": 741 }, { "epoch": 0.5650104702075005, "grad_norm": 39.43867111206055, "learning_rate": 8.329622986709418e-06, "loss": 0.9909, "step": 742 }, { "epoch": 0.5657719398438987, "grad_norm": 26.780221939086914, "learning_rate": 8.32507496088639e-06, "loss": 0.8134, "step": 743 }, { "epoch": 0.566533409480297, "grad_norm": 26.742109298706055, "learning_rate": 8.320521997242773e-06, "loss": 0.6863, "step": 744 }, { "epoch": 0.5672948791166952, "grad_norm": 33.564453125, "learning_rate": 8.315964102539835e-06, "loss": 0.8355, "step": 745 }, { "epoch": 0.5680563487530935, "grad_norm": 38.662662506103516, "learning_rate": 8.311401283546164e-06, "loss": 0.8981, "step": 746 }, { "epoch": 0.5688178183894917, "grad_norm": 34.034584045410156, "learning_rate": 8.306833547037662e-06, "loss": 0.7791, "step": 747 }, { "epoch": 0.56957928802589, "grad_norm": 33.56183624267578, "learning_rate": 8.302260899797532e-06, "loss": 0.6985, "step": 748 }, { "epoch": 0.5703407576622882, "grad_norm": 30.30634880065918, "learning_rate": 8.297683348616274e-06, "loss": 0.8649, "step": 749 }, { "epoch": 0.5711022272986864, "grad_norm": 30.258676528930664, "learning_rate": 8.293100900291667e-06, "loss": 0.6825, "step": 750 }, { "epoch": 0.5718636969350848, "grad_norm": 29.29175567626953, "learning_rate": 8.288513561628761e-06, "loss": 0.5259, "step": 751 }, { "epoch": 0.572625166571483, "grad_norm": 26.954097747802734, "learning_rate": 8.283921339439871e-06, "loss": 0.6971, "step": 752 }, { "epoch": 0.5733866362078812, "grad_norm": 26.251937866210938, "learning_rate": 8.279324240544565e-06, "loss": 0.6413, "step": 753 }, { "epoch": 0.5741481058442794, "grad_norm": 27.7598819732666, "learning_rate": 8.274722271769648e-06, "loss": 0.7426, "step": 754 }, { "epoch": 0.5749095754806777, "grad_norm": 38.244510650634766, "learning_rate": 8.270115439949164e-06, "loss": 0.82, "step": 755 }, { "epoch": 0.575671045117076, "grad_norm": 32.2285270690918, "learning_rate": 8.26550375192437e-06, "loss": 0.8009, "step": 756 }, { "epoch": 0.5764325147534742, "grad_norm": 34.027008056640625, "learning_rate": 8.260887214543747e-06, "loss": 0.8165, "step": 757 }, { "epoch": 0.5771939843898725, "grad_norm": 31.854063034057617, "learning_rate": 8.256265834662965e-06, "loss": 0.8268, "step": 758 }, { "epoch": 0.5779554540262707, "grad_norm": 34.228004455566406, "learning_rate": 8.251639619144888e-06, "loss": 0.7154, "step": 759 }, { "epoch": 0.5787169236626689, "grad_norm": 26.681699752807617, "learning_rate": 8.24700857485957e-06, "loss": 0.7981, "step": 760 }, { "epoch": 0.5794783932990673, "grad_norm": 30.64720916748047, "learning_rate": 8.242372708684223e-06, "loss": 0.8458, "step": 761 }, { "epoch": 0.5802398629354655, "grad_norm": 30.193696975708008, "learning_rate": 8.23773202750323e-06, "loss": 0.8717, "step": 762 }, { "epoch": 0.5810013325718637, "grad_norm": 29.077957153320312, "learning_rate": 8.233086538208115e-06, "loss": 0.7841, "step": 763 }, { "epoch": 0.5817628022082619, "grad_norm": 38.71321105957031, "learning_rate": 8.228436247697556e-06, "loss": 0.8796, "step": 764 }, { "epoch": 0.5825242718446602, "grad_norm": 31.869609832763672, "learning_rate": 8.223781162877341e-06, "loss": 0.8083, "step": 765 }, { "epoch": 0.5832857414810585, "grad_norm": 37.75756072998047, "learning_rate": 8.2191212906604e-06, "loss": 1.0693, "step": 766 }, { "epoch": 0.5840472111174567, "grad_norm": 34.645416259765625, "learning_rate": 8.214456637966754e-06, "loss": 0.9185, "step": 767 }, { "epoch": 0.584808680753855, "grad_norm": 27.102375030517578, "learning_rate": 8.209787211723534e-06, "loss": 0.659, "step": 768 }, { "epoch": 0.5855701503902532, "grad_norm": 27.539791107177734, "learning_rate": 8.205113018864954e-06, "loss": 0.6687, "step": 769 }, { "epoch": 0.5863316200266514, "grad_norm": 30.286901473999023, "learning_rate": 8.20043406633231e-06, "loss": 0.7485, "step": 770 }, { "epoch": 0.5870930896630497, "grad_norm": 33.38515090942383, "learning_rate": 8.195750361073966e-06, "loss": 0.8467, "step": 771 }, { "epoch": 0.587854559299448, "grad_norm": 44.452693939208984, "learning_rate": 8.191061910045341e-06, "loss": 0.6534, "step": 772 }, { "epoch": 0.5886160289358462, "grad_norm": 33.47657775878906, "learning_rate": 8.186368720208906e-06, "loss": 0.7487, "step": 773 }, { "epoch": 0.5893774985722444, "grad_norm": 29.39069938659668, "learning_rate": 8.181670798534164e-06, "loss": 0.8182, "step": 774 }, { "epoch": 0.5901389682086426, "grad_norm": 33.332374572753906, "learning_rate": 8.17696815199765e-06, "loss": 0.9018, "step": 775 }, { "epoch": 0.590900437845041, "grad_norm": 27.58820343017578, "learning_rate": 8.172260787582914e-06, "loss": 0.7479, "step": 776 }, { "epoch": 0.5916619074814392, "grad_norm": 31.04929542541504, "learning_rate": 8.167548712280507e-06, "loss": 0.7872, "step": 777 }, { "epoch": 0.5924233771178374, "grad_norm": 38.24930953979492, "learning_rate": 8.162831933087985e-06, "loss": 0.7874, "step": 778 }, { "epoch": 0.5931848467542357, "grad_norm": 36.27316665649414, "learning_rate": 8.158110457009882e-06, "loss": 0.8672, "step": 779 }, { "epoch": 0.5939463163906339, "grad_norm": 36.5713996887207, "learning_rate": 8.15338429105771e-06, "loss": 0.5997, "step": 780 }, { "epoch": 0.5947077860270322, "grad_norm": 35.257774353027344, "learning_rate": 8.148653442249949e-06, "loss": 0.8171, "step": 781 }, { "epoch": 0.5954692556634305, "grad_norm": 36.267547607421875, "learning_rate": 8.143917917612023e-06, "loss": 0.8415, "step": 782 }, { "epoch": 0.5962307252998287, "grad_norm": 27.755474090576172, "learning_rate": 8.13917772417631e-06, "loss": 0.6817, "step": 783 }, { "epoch": 0.5969921949362269, "grad_norm": 36.79349136352539, "learning_rate": 8.134432868982118e-06, "loss": 0.7736, "step": 784 }, { "epoch": 0.5977536645726251, "grad_norm": 32.57246398925781, "learning_rate": 8.129683359075675e-06, "loss": 0.7669, "step": 785 }, { "epoch": 0.5985151342090235, "grad_norm": 37.34968185424805, "learning_rate": 8.124929201510124e-06, "loss": 0.7779, "step": 786 }, { "epoch": 0.5992766038454217, "grad_norm": 40.94609069824219, "learning_rate": 8.12017040334551e-06, "loss": 0.9911, "step": 787 }, { "epoch": 0.6000380734818199, "grad_norm": 28.85545539855957, "learning_rate": 8.11540697164877e-06, "loss": 0.6715, "step": 788 }, { "epoch": 0.6007995431182181, "grad_norm": 49.465614318847656, "learning_rate": 8.110638913493718e-06, "loss": 0.6255, "step": 789 }, { "epoch": 0.6015610127546164, "grad_norm": 50.699974060058594, "learning_rate": 8.105866235961045e-06, "loss": 0.566, "step": 790 }, { "epoch": 0.6023224823910147, "grad_norm": 27.396591186523438, "learning_rate": 8.101088946138294e-06, "loss": 0.6544, "step": 791 }, { "epoch": 0.6030839520274129, "grad_norm": 32.85529708862305, "learning_rate": 8.096307051119863e-06, "loss": 0.8224, "step": 792 }, { "epoch": 0.6038454216638112, "grad_norm": 26.31869125366211, "learning_rate": 8.091520558006988e-06, "loss": 0.7248, "step": 793 }, { "epoch": 0.6046068913002094, "grad_norm": 28.539140701293945, "learning_rate": 8.086729473907734e-06, "loss": 0.7012, "step": 794 }, { "epoch": 0.6053683609366076, "grad_norm": 34.67433166503906, "learning_rate": 8.081933805936977e-06, "loss": 0.793, "step": 795 }, { "epoch": 0.606129830573006, "grad_norm": 26.926082611083984, "learning_rate": 8.07713356121641e-06, "loss": 0.6892, "step": 796 }, { "epoch": 0.6068913002094042, "grad_norm": 37.93033981323242, "learning_rate": 8.07232874687452e-06, "loss": 0.7574, "step": 797 }, { "epoch": 0.6076527698458024, "grad_norm": 61.16469192504883, "learning_rate": 8.067519370046572e-06, "loss": 1.1853, "step": 798 }, { "epoch": 0.6084142394822006, "grad_norm": 36.617897033691406, "learning_rate": 8.06270543787462e-06, "loss": 0.9867, "step": 799 }, { "epoch": 0.6091757091185989, "grad_norm": 41.403324127197266, "learning_rate": 8.057886957507466e-06, "loss": 1.0133, "step": 800 }, { "epoch": 0.6099371787549971, "grad_norm": 35.976966857910156, "learning_rate": 8.053063936100682e-06, "loss": 0.9601, "step": 801 }, { "epoch": 0.6106986483913954, "grad_norm": 28.360689163208008, "learning_rate": 8.048236380816576e-06, "loss": 0.7825, "step": 802 }, { "epoch": 0.6114601180277937, "grad_norm": 28.78184700012207, "learning_rate": 8.043404298824189e-06, "loss": 0.7889, "step": 803 }, { "epoch": 0.6122215876641919, "grad_norm": 21.7941951751709, "learning_rate": 8.038567697299287e-06, "loss": 0.5789, "step": 804 }, { "epoch": 0.6129830573005901, "grad_norm": 33.62324142456055, "learning_rate": 8.033726583424341e-06, "loss": 0.8517, "step": 805 }, { "epoch": 0.6137445269369883, "grad_norm": 33.0076789855957, "learning_rate": 8.028880964388534e-06, "loss": 0.8585, "step": 806 }, { "epoch": 0.6145059965733867, "grad_norm": 27.630537033081055, "learning_rate": 8.024030847387729e-06, "loss": 0.7323, "step": 807 }, { "epoch": 0.6152674662097849, "grad_norm": 24.592273712158203, "learning_rate": 8.019176239624476e-06, "loss": 0.7206, "step": 808 }, { "epoch": 0.6160289358461831, "grad_norm": 26.27126693725586, "learning_rate": 8.014317148307987e-06, "loss": 0.7633, "step": 809 }, { "epoch": 0.6167904054825813, "grad_norm": 28.837806701660156, "learning_rate": 8.009453580654138e-06, "loss": 0.5625, "step": 810 }, { "epoch": 0.6175518751189796, "grad_norm": 22.987646102905273, "learning_rate": 8.004585543885448e-06, "loss": 0.6402, "step": 811 }, { "epoch": 0.6183133447553779, "grad_norm": 26.343568801879883, "learning_rate": 7.999713045231078e-06, "loss": 0.6907, "step": 812 }, { "epoch": 0.6190748143917761, "grad_norm": 25.834089279174805, "learning_rate": 7.99483609192681e-06, "loss": 0.5879, "step": 813 }, { "epoch": 0.6198362840281744, "grad_norm": 26.641136169433594, "learning_rate": 7.989954691215044e-06, "loss": 0.7393, "step": 814 }, { "epoch": 0.6205977536645726, "grad_norm": 31.973413467407227, "learning_rate": 7.985068850344783e-06, "loss": 0.8564, "step": 815 }, { "epoch": 0.6213592233009708, "grad_norm": 32.119537353515625, "learning_rate": 7.980178576571626e-06, "loss": 0.9437, "step": 816 }, { "epoch": 0.6221206929373692, "grad_norm": 25.91680908203125, "learning_rate": 7.975283877157752e-06, "loss": 0.6618, "step": 817 }, { "epoch": 0.6228821625737674, "grad_norm": 26.057126998901367, "learning_rate": 7.970384759371914e-06, "loss": 0.7615, "step": 818 }, { "epoch": 0.6236436322101656, "grad_norm": 28.24910545349121, "learning_rate": 7.965481230489425e-06, "loss": 0.7863, "step": 819 }, { "epoch": 0.6244051018465638, "grad_norm": 32.42158889770508, "learning_rate": 7.960573297792153e-06, "loss": 0.7045, "step": 820 }, { "epoch": 0.6251665714829621, "grad_norm": 33.47333526611328, "learning_rate": 7.955660968568499e-06, "loss": 0.8996, "step": 821 }, { "epoch": 0.6259280411193604, "grad_norm": 33.64387512207031, "learning_rate": 7.950744250113398e-06, "loss": 0.8674, "step": 822 }, { "epoch": 0.6266895107557586, "grad_norm": 21.23735809326172, "learning_rate": 7.945823149728304e-06, "loss": 0.6238, "step": 823 }, { "epoch": 0.6274509803921569, "grad_norm": 33.83268356323242, "learning_rate": 7.94089767472117e-06, "loss": 0.7337, "step": 824 }, { "epoch": 0.6282124500285551, "grad_norm": 31.84981918334961, "learning_rate": 7.935967832406456e-06, "loss": 0.7967, "step": 825 }, { "epoch": 0.6289739196649533, "grad_norm": 31.125083923339844, "learning_rate": 7.9310336301051e-06, "loss": 1.0462, "step": 826 }, { "epoch": 0.6297353893013516, "grad_norm": 26.396533966064453, "learning_rate": 7.92609507514452e-06, "loss": 0.6512, "step": 827 }, { "epoch": 0.6304968589377499, "grad_norm": 26.503463745117188, "learning_rate": 7.921152174858592e-06, "loss": 0.7569, "step": 828 }, { "epoch": 0.6312583285741481, "grad_norm": 34.26448440551758, "learning_rate": 7.91620493658765e-06, "loss": 0.799, "step": 829 }, { "epoch": 0.6320197982105463, "grad_norm": 32.36408233642578, "learning_rate": 7.911253367678469e-06, "loss": 0.9017, "step": 830 }, { "epoch": 0.6327812678469446, "grad_norm": 30.752483367919922, "learning_rate": 7.906297475484255e-06, "loss": 0.8698, "step": 831 }, { "epoch": 0.6335427374833429, "grad_norm": 27.395137786865234, "learning_rate": 7.90133726736463e-06, "loss": 0.6505, "step": 832 }, { "epoch": 0.6343042071197411, "grad_norm": 32.6951789855957, "learning_rate": 7.896372750685632e-06, "loss": 0.8879, "step": 833 }, { "epoch": 0.6350656767561393, "grad_norm": 39.803714752197266, "learning_rate": 7.891403932819692e-06, "loss": 0.9498, "step": 834 }, { "epoch": 0.6358271463925376, "grad_norm": 34.31433868408203, "learning_rate": 7.886430821145633e-06, "loss": 0.9549, "step": 835 }, { "epoch": 0.6365886160289358, "grad_norm": 27.380159378051758, "learning_rate": 7.881453423048651e-06, "loss": 0.7375, "step": 836 }, { "epoch": 0.6373500856653341, "grad_norm": 26.90700912475586, "learning_rate": 7.876471745920304e-06, "loss": 0.6806, "step": 837 }, { "epoch": 0.6381115553017324, "grad_norm": 25.832687377929688, "learning_rate": 7.871485797158515e-06, "loss": 0.8036, "step": 838 }, { "epoch": 0.6388730249381306, "grad_norm": 34.30730056762695, "learning_rate": 7.866495584167542e-06, "loss": 0.9942, "step": 839 }, { "epoch": 0.6396344945745288, "grad_norm": 25.98689079284668, "learning_rate": 7.861501114357976e-06, "loss": 0.7233, "step": 840 }, { "epoch": 0.640395964210927, "grad_norm": 32.80167007446289, "learning_rate": 7.856502395146735e-06, "loss": 0.7946, "step": 841 }, { "epoch": 0.6411574338473254, "grad_norm": 34.85222625732422, "learning_rate": 7.851499433957039e-06, "loss": 1.0334, "step": 842 }, { "epoch": 0.6419189034837236, "grad_norm": 36.892765045166016, "learning_rate": 7.846492238218414e-06, "loss": 0.8306, "step": 843 }, { "epoch": 0.6426803731201218, "grad_norm": 29.023441314697266, "learning_rate": 7.841480815366676e-06, "loss": 0.9039, "step": 844 }, { "epoch": 0.64344184275652, "grad_norm": 30.305448532104492, "learning_rate": 7.83646517284391e-06, "loss": 0.6917, "step": 845 }, { "epoch": 0.6442033123929183, "grad_norm": 32.46067428588867, "learning_rate": 7.831445318098476e-06, "loss": 0.7484, "step": 846 }, { "epoch": 0.6449647820293166, "grad_norm": 29.135446548461914, "learning_rate": 7.826421258584984e-06, "loss": 0.7558, "step": 847 }, { "epoch": 0.6457262516657148, "grad_norm": 27.77723503112793, "learning_rate": 7.821393001764287e-06, "loss": 0.7524, "step": 848 }, { "epoch": 0.6464877213021131, "grad_norm": 26.93240737915039, "learning_rate": 7.81636055510348e-06, "loss": 0.8165, "step": 849 }, { "epoch": 0.6472491909385113, "grad_norm": 29.845855712890625, "learning_rate": 7.811323926075867e-06, "loss": 0.7827, "step": 850 }, { "epoch": 0.6480106605749095, "grad_norm": 28.056304931640625, "learning_rate": 7.806283122160974e-06, "loss": 0.7516, "step": 851 }, { "epoch": 0.6487721302113079, "grad_norm": 40.83567428588867, "learning_rate": 7.801238150844523e-06, "loss": 0.8165, "step": 852 }, { "epoch": 0.6495335998477061, "grad_norm": 26.75080680847168, "learning_rate": 7.796189019618421e-06, "loss": 0.5547, "step": 853 }, { "epoch": 0.6502950694841043, "grad_norm": 35.45816421508789, "learning_rate": 7.791135735980758e-06, "loss": 0.9976, "step": 854 }, { "epoch": 0.6510565391205025, "grad_norm": 24.48919105529785, "learning_rate": 7.786078307435788e-06, "loss": 0.5694, "step": 855 }, { "epoch": 0.6518180087569008, "grad_norm": 25.51799964904785, "learning_rate": 7.781016741493922e-06, "loss": 0.7217, "step": 856 }, { "epoch": 0.6525794783932991, "grad_norm": 52.26631164550781, "learning_rate": 7.77595104567171e-06, "loss": 0.7708, "step": 857 }, { "epoch": 0.6533409480296973, "grad_norm": 27.07032012939453, "learning_rate": 7.770881227491842e-06, "loss": 0.7683, "step": 858 }, { "epoch": 0.6541024176660956, "grad_norm": 57.096466064453125, "learning_rate": 7.765807294483125e-06, "loss": 0.7533, "step": 859 }, { "epoch": 0.6548638873024938, "grad_norm": 30.253202438354492, "learning_rate": 7.76072925418048e-06, "loss": 0.8218, "step": 860 }, { "epoch": 0.655625356938892, "grad_norm": 31.545814514160156, "learning_rate": 7.755647114124922e-06, "loss": 0.8249, "step": 861 }, { "epoch": 0.6563868265752903, "grad_norm": 28.886499404907227, "learning_rate": 7.750560881863563e-06, "loss": 0.8746, "step": 862 }, { "epoch": 0.6571482962116886, "grad_norm": 26.418315887451172, "learning_rate": 7.74547056494958e-06, "loss": 0.6539, "step": 863 }, { "epoch": 0.6579097658480868, "grad_norm": 31.141759872436523, "learning_rate": 7.740376170942227e-06, "loss": 0.9195, "step": 864 }, { "epoch": 0.658671235484485, "grad_norm": 29.36141014099121, "learning_rate": 7.73527770740681e-06, "loss": 0.886, "step": 865 }, { "epoch": 0.6594327051208833, "grad_norm": 27.355432510375977, "learning_rate": 7.730175181914672e-06, "loss": 0.7236, "step": 866 }, { "epoch": 0.6601941747572816, "grad_norm": 45.42149353027344, "learning_rate": 7.725068602043195e-06, "loss": 1.1618, "step": 867 }, { "epoch": 0.6609556443936798, "grad_norm": 33.360931396484375, "learning_rate": 7.719957975375779e-06, "loss": 0.8988, "step": 868 }, { "epoch": 0.661717114030078, "grad_norm": 28.71207046508789, "learning_rate": 7.714843309501836e-06, "loss": 0.5699, "step": 869 }, { "epoch": 0.6624785836664763, "grad_norm": 40.582035064697266, "learning_rate": 7.709724612016772e-06, "loss": 0.8159, "step": 870 }, { "epoch": 0.6632400533028745, "grad_norm": 40.518150329589844, "learning_rate": 7.704601890521982e-06, "loss": 0.8031, "step": 871 }, { "epoch": 0.6640015229392728, "grad_norm": 32.10520553588867, "learning_rate": 7.699475152624839e-06, "loss": 0.7844, "step": 872 }, { "epoch": 0.6647629925756711, "grad_norm": 33.51770782470703, "learning_rate": 7.694344405938679e-06, "loss": 0.9458, "step": 873 }, { "epoch": 0.6655244622120693, "grad_norm": 28.527631759643555, "learning_rate": 7.68920965808279e-06, "loss": 0.8622, "step": 874 }, { "epoch": 0.6662859318484675, "grad_norm": 38.19065856933594, "learning_rate": 7.684070916682403e-06, "loss": 0.6396, "step": 875 }, { "epoch": 0.6670474014848657, "grad_norm": 27.93634033203125, "learning_rate": 7.67892818936868e-06, "loss": 0.7168, "step": 876 }, { "epoch": 0.6678088711212641, "grad_norm": 28.99605941772461, "learning_rate": 7.673781483778698e-06, "loss": 0.7734, "step": 877 }, { "epoch": 0.6685703407576623, "grad_norm": 39.4337158203125, "learning_rate": 7.66863080755545e-06, "loss": 1.4895, "step": 878 }, { "epoch": 0.6693318103940605, "grad_norm": 28.5806827545166, "learning_rate": 7.663476168347818e-06, "loss": 0.7658, "step": 879 }, { "epoch": 0.6700932800304588, "grad_norm": 37.91471481323242, "learning_rate": 7.658317573810572e-06, "loss": 0.8842, "step": 880 }, { "epoch": 0.670854749666857, "grad_norm": 31.70551872253418, "learning_rate": 7.653155031604356e-06, "loss": 0.7191, "step": 881 }, { "epoch": 0.6716162193032553, "grad_norm": 31.07270050048828, "learning_rate": 7.647988549395676e-06, "loss": 0.8618, "step": 882 }, { "epoch": 0.6723776889396536, "grad_norm": 32.019893646240234, "learning_rate": 7.642818134856888e-06, "loss": 0.7914, "step": 883 }, { "epoch": 0.6731391585760518, "grad_norm": 30.21106719970703, "learning_rate": 7.637643795666192e-06, "loss": 0.6793, "step": 884 }, { "epoch": 0.67390062821245, "grad_norm": 23.416296005249023, "learning_rate": 7.632465539507608e-06, "loss": 0.6975, "step": 885 }, { "epoch": 0.6746620978488482, "grad_norm": 28.055620193481445, "learning_rate": 7.62728337407098e-06, "loss": 0.827, "step": 886 }, { "epoch": 0.6754235674852466, "grad_norm": 24.000978469848633, "learning_rate": 7.622097307051957e-06, "loss": 0.5859, "step": 887 }, { "epoch": 0.6761850371216448, "grad_norm": 29.766515731811523, "learning_rate": 7.616907346151977e-06, "loss": 0.9002, "step": 888 }, { "epoch": 0.676946506758043, "grad_norm": 28.786598205566406, "learning_rate": 7.611713499078265e-06, "loss": 0.8271, "step": 889 }, { "epoch": 0.6777079763944412, "grad_norm": 30.589488983154297, "learning_rate": 7.606515773543814e-06, "loss": 0.6112, "step": 890 }, { "epoch": 0.6784694460308395, "grad_norm": 24.776464462280273, "learning_rate": 7.601314177267381e-06, "loss": 0.7205, "step": 891 }, { "epoch": 0.6792309156672378, "grad_norm": 26.0139217376709, "learning_rate": 7.596108717973465e-06, "loss": 0.7979, "step": 892 }, { "epoch": 0.679992385303636, "grad_norm": 66.2892074584961, "learning_rate": 7.590899403392307e-06, "loss": 0.7666, "step": 893 }, { "epoch": 0.6807538549400343, "grad_norm": 30.251981735229492, "learning_rate": 7.585686241259871e-06, "loss": 0.7796, "step": 894 }, { "epoch": 0.6815153245764325, "grad_norm": 30.38053321838379, "learning_rate": 7.5804692393178334e-06, "loss": 0.8498, "step": 895 }, { "epoch": 0.6822767942128307, "grad_norm": 34.0738639831543, "learning_rate": 7.575248405313578e-06, "loss": 0.9128, "step": 896 }, { "epoch": 0.683038263849229, "grad_norm": 28.353290557861328, "learning_rate": 7.5700237470001705e-06, "loss": 0.7225, "step": 897 }, { "epoch": 0.6837997334856273, "grad_norm": 25.537622451782227, "learning_rate": 7.564795272136365e-06, "loss": 0.6939, "step": 898 }, { "epoch": 0.6845612031220255, "grad_norm": 28.82963752746582, "learning_rate": 7.559562988486576e-06, "loss": 0.6956, "step": 899 }, { "epoch": 0.6853226727584237, "grad_norm": 24.72161865234375, "learning_rate": 7.554326903820881e-06, "loss": 0.7923, "step": 900 }, { "epoch": 0.686084142394822, "grad_norm": 33.14993667602539, "learning_rate": 7.549087025914995e-06, "loss": 0.8066, "step": 901 }, { "epoch": 0.6868456120312203, "grad_norm": 35.39863204956055, "learning_rate": 7.5438433625502725e-06, "loss": 0.9213, "step": 902 }, { "epoch": 0.6876070816676185, "grad_norm": 29.69724464416504, "learning_rate": 7.538595921513683e-06, "loss": 0.7398, "step": 903 }, { "epoch": 0.6883685513040168, "grad_norm": 35.563053131103516, "learning_rate": 7.533344710597814e-06, "loss": 0.7713, "step": 904 }, { "epoch": 0.689130020940415, "grad_norm": 30.227357864379883, "learning_rate": 7.5280897376008415e-06, "loss": 0.7612, "step": 905 }, { "epoch": 0.6898914905768132, "grad_norm": 27.88680076599121, "learning_rate": 7.522831010326538e-06, "loss": 0.7491, "step": 906 }, { "epoch": 0.6906529602132115, "grad_norm": 37.14356994628906, "learning_rate": 7.5175685365842435e-06, "loss": 0.9955, "step": 907 }, { "epoch": 0.6914144298496098, "grad_norm": 28.307918548583984, "learning_rate": 7.512302324188867e-06, "loss": 0.8756, "step": 908 }, { "epoch": 0.692175899486008, "grad_norm": 36.45597457885742, "learning_rate": 7.507032380960868e-06, "loss": 0.8675, "step": 909 }, { "epoch": 0.6929373691224062, "grad_norm": 34.10755157470703, "learning_rate": 7.5017587147262425e-06, "loss": 1.0245, "step": 910 }, { "epoch": 0.6936988387588044, "grad_norm": 33.00735092163086, "learning_rate": 7.496481333316522e-06, "loss": 0.7733, "step": 911 }, { "epoch": 0.6944603083952028, "grad_norm": 36.1781005859375, "learning_rate": 7.491200244568748e-06, "loss": 0.9425, "step": 912 }, { "epoch": 0.695221778031601, "grad_norm": 25.176944732666016, "learning_rate": 7.4859154563254766e-06, "loss": 0.7413, "step": 913 }, { "epoch": 0.6959832476679992, "grad_norm": 29.096973419189453, "learning_rate": 7.480626976434746e-06, "loss": 0.7264, "step": 914 }, { "epoch": 0.6967447173043975, "grad_norm": 28.575490951538086, "learning_rate": 7.475334812750089e-06, "loss": 0.7541, "step": 915 }, { "epoch": 0.6975061869407957, "grad_norm": 28.50647735595703, "learning_rate": 7.470038973130497e-06, "loss": 0.6062, "step": 916 }, { "epoch": 0.698267656577194, "grad_norm": 30.17618179321289, "learning_rate": 7.464739465440431e-06, "loss": 0.8535, "step": 917 }, { "epoch": 0.6990291262135923, "grad_norm": 21.86348533630371, "learning_rate": 7.4594362975497896e-06, "loss": 0.5617, "step": 918 }, { "epoch": 0.6997905958499905, "grad_norm": 23.84324073791504, "learning_rate": 7.454129477333916e-06, "loss": 0.6717, "step": 919 }, { "epoch": 0.7005520654863887, "grad_norm": 29.712072372436523, "learning_rate": 7.44881901267357e-06, "loss": 0.9733, "step": 920 }, { "epoch": 0.7013135351227869, "grad_norm": 27.97547149658203, "learning_rate": 7.443504911454927e-06, "loss": 0.7411, "step": 921 }, { "epoch": 0.7020750047591853, "grad_norm": 32.59486770629883, "learning_rate": 7.43818718156956e-06, "loss": 0.8966, "step": 922 }, { "epoch": 0.7028364743955835, "grad_norm": 31.34674644470215, "learning_rate": 7.432865830914435e-06, "loss": 0.9788, "step": 923 }, { "epoch": 0.7035979440319817, "grad_norm": 18.920564651489258, "learning_rate": 7.427540867391894e-06, "loss": 0.6062, "step": 924 }, { "epoch": 0.70435941366838, "grad_norm": 21.33208465576172, "learning_rate": 7.422212298909638e-06, "loss": 0.6215, "step": 925 }, { "epoch": 0.7051208833047782, "grad_norm": 32.146976470947266, "learning_rate": 7.416880133380732e-06, "loss": 0.8168, "step": 926 }, { "epoch": 0.7058823529411765, "grad_norm": 25.904571533203125, "learning_rate": 7.411544378723572e-06, "loss": 0.7482, "step": 927 }, { "epoch": 0.7066438225775747, "grad_norm": 29.501108169555664, "learning_rate": 7.406205042861892e-06, "loss": 0.778, "step": 928 }, { "epoch": 0.707405292213973, "grad_norm": 38.986534118652344, "learning_rate": 7.400862133724739e-06, "loss": 0.891, "step": 929 }, { "epoch": 0.7081667618503712, "grad_norm": 24.509849548339844, "learning_rate": 7.39551565924647e-06, "loss": 0.7243, "step": 930 }, { "epoch": 0.7089282314867694, "grad_norm": 24.55140495300293, "learning_rate": 7.390165627366735e-06, "loss": 0.7393, "step": 931 }, { "epoch": 0.7096897011231678, "grad_norm": 28.782194137573242, "learning_rate": 7.384812046030468e-06, "loss": 0.7491, "step": 932 }, { "epoch": 0.710451170759566, "grad_norm": 26.1166934967041, "learning_rate": 7.379454923187868e-06, "loss": 0.667, "step": 933 }, { "epoch": 0.7112126403959642, "grad_norm": 30.481565475463867, "learning_rate": 7.374094266794403e-06, "loss": 0.8279, "step": 934 }, { "epoch": 0.7119741100323624, "grad_norm": 28.972707748413086, "learning_rate": 7.368730084810782e-06, "loss": 0.6376, "step": 935 }, { "epoch": 0.7127355796687607, "grad_norm": 26.157737731933594, "learning_rate": 7.363362385202952e-06, "loss": 0.8613, "step": 936 }, { "epoch": 0.713497049305159, "grad_norm": 26.83064842224121, "learning_rate": 7.357991175942083e-06, "loss": 0.777, "step": 937 }, { "epoch": 0.7142585189415572, "grad_norm": 26.857431411743164, "learning_rate": 7.352616465004556e-06, "loss": 0.6033, "step": 938 }, { "epoch": 0.7150199885779555, "grad_norm": 28.428842544555664, "learning_rate": 7.347238260371954e-06, "loss": 0.9263, "step": 939 }, { "epoch": 0.7157814582143537, "grad_norm": 28.889423370361328, "learning_rate": 7.341856570031045e-06, "loss": 0.6927, "step": 940 }, { "epoch": 0.7165429278507519, "grad_norm": 33.6306266784668, "learning_rate": 7.336471401973778e-06, "loss": 0.7787, "step": 941 }, { "epoch": 0.7173043974871502, "grad_norm": 23.571929931640625, "learning_rate": 7.331082764197262e-06, "loss": 0.565, "step": 942 }, { "epoch": 0.7180658671235485, "grad_norm": 29.913192749023438, "learning_rate": 7.325690664703763e-06, "loss": 0.8598, "step": 943 }, { "epoch": 0.7188273367599467, "grad_norm": 30.621116638183594, "learning_rate": 7.320295111500683e-06, "loss": 0.6628, "step": 944 }, { "epoch": 0.7195888063963449, "grad_norm": 35.32845687866211, "learning_rate": 7.314896112600558e-06, "loss": 0.9929, "step": 945 }, { "epoch": 0.7203502760327432, "grad_norm": 27.44571304321289, "learning_rate": 7.309493676021035e-06, "loss": 0.8326, "step": 946 }, { "epoch": 0.7211117456691415, "grad_norm": 32.142967224121094, "learning_rate": 7.30408780978487e-06, "loss": 0.8856, "step": 947 }, { "epoch": 0.7218732153055397, "grad_norm": 30.383384704589844, "learning_rate": 7.298678521919914e-06, "loss": 0.7051, "step": 948 }, { "epoch": 0.7226346849419379, "grad_norm": 29.19371795654297, "learning_rate": 7.293265820459095e-06, "loss": 0.8118, "step": 949 }, { "epoch": 0.7233961545783362, "grad_norm": 31.852752685546875, "learning_rate": 7.28784971344041e-06, "loss": 0.7124, "step": 950 }, { "epoch": 0.7241576242147344, "grad_norm": 26.42278289794922, "learning_rate": 7.2824302089069184e-06, "loss": 0.7328, "step": 951 }, { "epoch": 0.7249190938511327, "grad_norm": 31.66285514831543, "learning_rate": 7.277007314906719e-06, "loss": 0.814, "step": 952 }, { "epoch": 0.725680563487531, "grad_norm": 26.553312301635742, "learning_rate": 7.271581039492949e-06, "loss": 0.7502, "step": 953 }, { "epoch": 0.7264420331239292, "grad_norm": 36.027652740478516, "learning_rate": 7.2661513907237634e-06, "loss": 1.0965, "step": 954 }, { "epoch": 0.7272035027603274, "grad_norm": 25.32257652282715, "learning_rate": 7.260718376662329e-06, "loss": 0.8641, "step": 955 }, { "epoch": 0.7279649723967256, "grad_norm": 32.486141204833984, "learning_rate": 7.25528200537681e-06, "loss": 0.8929, "step": 956 }, { "epoch": 0.728726442033124, "grad_norm": 46.99209976196289, "learning_rate": 7.249842284940352e-06, "loss": 0.8834, "step": 957 }, { "epoch": 0.7294879116695222, "grad_norm": 25.190576553344727, "learning_rate": 7.244399223431082e-06, "loss": 0.6918, "step": 958 }, { "epoch": 0.7302493813059204, "grad_norm": 28.75440788269043, "learning_rate": 7.238952828932084e-06, "loss": 0.8444, "step": 959 }, { "epoch": 0.7310108509423187, "grad_norm": 26.397907257080078, "learning_rate": 7.233503109531389e-06, "loss": 0.6744, "step": 960 }, { "epoch": 0.7317723205787169, "grad_norm": 34.44126892089844, "learning_rate": 7.228050073321971e-06, "loss": 0.8709, "step": 961 }, { "epoch": 0.7325337902151152, "grad_norm": 28.410762786865234, "learning_rate": 7.222593728401724e-06, "loss": 0.8122, "step": 962 }, { "epoch": 0.7332952598515134, "grad_norm": 30.714998245239258, "learning_rate": 7.2171340828734626e-06, "loss": 0.7721, "step": 963 }, { "epoch": 0.7340567294879117, "grad_norm": 24.129196166992188, "learning_rate": 7.211671144844896e-06, "loss": 0.5979, "step": 964 }, { "epoch": 0.7348181991243099, "grad_norm": 36.182044982910156, "learning_rate": 7.206204922428625e-06, "loss": 0.9329, "step": 965 }, { "epoch": 0.7355796687607081, "grad_norm": 28.672395706176758, "learning_rate": 7.200735423742131e-06, "loss": 0.7124, "step": 966 }, { "epoch": 0.7363411383971065, "grad_norm": 23.973031997680664, "learning_rate": 7.195262656907758e-06, "loss": 0.7238, "step": 967 }, { "epoch": 0.7371026080335047, "grad_norm": 26.83302116394043, "learning_rate": 7.1897866300527e-06, "loss": 0.8061, "step": 968 }, { "epoch": 0.7378640776699029, "grad_norm": 24.463834762573242, "learning_rate": 7.1843073513089985e-06, "loss": 0.6155, "step": 969 }, { "epoch": 0.7386255473063011, "grad_norm": 27.022062301635742, "learning_rate": 7.178824828813522e-06, "loss": 0.6644, "step": 970 }, { "epoch": 0.7393870169426994, "grad_norm": 28.182111740112305, "learning_rate": 7.173339070707953e-06, "loss": 0.8427, "step": 971 }, { "epoch": 0.7401484865790977, "grad_norm": 33.68251037597656, "learning_rate": 7.1678500851387835e-06, "loss": 0.873, "step": 972 }, { "epoch": 0.7409099562154959, "grad_norm": 28.491044998168945, "learning_rate": 7.162357880257294e-06, "loss": 0.8262, "step": 973 }, { "epoch": 0.7416714258518942, "grad_norm": 28.93327522277832, "learning_rate": 7.1568624642195495e-06, "loss": 0.7282, "step": 974 }, { "epoch": 0.7424328954882924, "grad_norm": 22.979516983032227, "learning_rate": 7.151363845186382e-06, "loss": 0.5553, "step": 975 }, { "epoch": 0.7431943651246906, "grad_norm": 28.500520706176758, "learning_rate": 7.14586203132338e-06, "loss": 0.8657, "step": 976 }, { "epoch": 0.743955834761089, "grad_norm": 38.0922966003418, "learning_rate": 7.140357030800875e-06, "loss": 0.9531, "step": 977 }, { "epoch": 0.7447173043974872, "grad_norm": 30.303611755371094, "learning_rate": 7.134848851793934e-06, "loss": 0.7469, "step": 978 }, { "epoch": 0.7454787740338854, "grad_norm": 34.94667053222656, "learning_rate": 7.129337502482343e-06, "loss": 0.9157, "step": 979 }, { "epoch": 0.7462402436702836, "grad_norm": 28.89977264404297, "learning_rate": 7.123822991050594e-06, "loss": 0.7208, "step": 980 }, { "epoch": 0.7470017133066819, "grad_norm": 28.439899444580078, "learning_rate": 7.118305325687876e-06, "loss": 0.7317, "step": 981 }, { "epoch": 0.7477631829430802, "grad_norm": 26.193099975585938, "learning_rate": 7.112784514588064e-06, "loss": 0.6008, "step": 982 }, { "epoch": 0.7485246525794784, "grad_norm": 25.30535125732422, "learning_rate": 7.1072605659497e-06, "loss": 0.754, "step": 983 }, { "epoch": 0.7492861222158766, "grad_norm": 23.79256248474121, "learning_rate": 7.101733487975988e-06, "loss": 0.6873, "step": 984 }, { "epoch": 0.7500475918522749, "grad_norm": 26.857532501220703, "learning_rate": 7.096203288874781e-06, "loss": 0.9162, "step": 985 }, { "epoch": 0.7508090614886731, "grad_norm": 31.37613296508789, "learning_rate": 7.090669976858563e-06, "loss": 0.7478, "step": 986 }, { "epoch": 0.7515705311250714, "grad_norm": 26.164493560791016, "learning_rate": 7.085133560144443e-06, "loss": 0.7949, "step": 987 }, { "epoch": 0.7523320007614697, "grad_norm": 31.36321449279785, "learning_rate": 7.07959404695414e-06, "loss": 0.751, "step": 988 }, { "epoch": 0.7530934703978679, "grad_norm": 27.06949806213379, "learning_rate": 7.074051445513972e-06, "loss": 0.7524, "step": 989 }, { "epoch": 0.7538549400342661, "grad_norm": 33.156089782714844, "learning_rate": 7.0685057640548415e-06, "loss": 0.838, "step": 990 }, { "epoch": 0.7546164096706643, "grad_norm": 37.368370056152344, "learning_rate": 7.062957010812227e-06, "loss": 0.8547, "step": 991 }, { "epoch": 0.7553778793070627, "grad_norm": 32.58074951171875, "learning_rate": 7.057405194026168e-06, "loss": 0.6642, "step": 992 }, { "epoch": 0.7561393489434609, "grad_norm": 28.085039138793945, "learning_rate": 7.051850321941252e-06, "loss": 0.7924, "step": 993 }, { "epoch": 0.7569008185798591, "grad_norm": 26.330997467041016, "learning_rate": 7.046292402806605e-06, "loss": 0.7999, "step": 994 }, { "epoch": 0.7576622882162574, "grad_norm": 38.67786407470703, "learning_rate": 7.040731444875878e-06, "loss": 1.0057, "step": 995 }, { "epoch": 0.7584237578526556, "grad_norm": 25.687808990478516, "learning_rate": 7.035167456407237e-06, "loss": 0.712, "step": 996 }, { "epoch": 0.7591852274890539, "grad_norm": 30.188947677612305, "learning_rate": 7.029600445663341e-06, "loss": 0.7298, "step": 997 }, { "epoch": 0.7599466971254522, "grad_norm": 26.661039352416992, "learning_rate": 7.024030420911346e-06, "loss": 0.8296, "step": 998 }, { "epoch": 0.7607081667618504, "grad_norm": 31.66457176208496, "learning_rate": 7.018457390422878e-06, "loss": 0.84, "step": 999 }, { "epoch": 0.7614696363982486, "grad_norm": 26.696500778198242, "learning_rate": 7.012881362474029e-06, "loss": 0.8177, "step": 1000 }, { "epoch": 0.7622311060346468, "grad_norm": 30.383148193359375, "learning_rate": 7.0073023453453414e-06, "loss": 0.9013, "step": 1001 }, { "epoch": 0.7629925756710452, "grad_norm": 31.11835289001465, "learning_rate": 7.0017203473217984e-06, "loss": 0.8432, "step": 1002 }, { "epoch": 0.7637540453074434, "grad_norm": 31.29349136352539, "learning_rate": 6.996135376692807e-06, "loss": 0.5571, "step": 1003 }, { "epoch": 0.7645155149438416, "grad_norm": 33.65591812133789, "learning_rate": 6.9905474417521905e-06, "loss": 0.74, "step": 1004 }, { "epoch": 0.7652769845802398, "grad_norm": 33.48772048950195, "learning_rate": 6.9849565507981754e-06, "loss": 0.6417, "step": 1005 }, { "epoch": 0.7660384542166381, "grad_norm": 34.43278884887695, "learning_rate": 6.979362712133373e-06, "loss": 0.8469, "step": 1006 }, { "epoch": 0.7667999238530364, "grad_norm": 30.405088424682617, "learning_rate": 6.97376593406478e-06, "loss": 0.8019, "step": 1007 }, { "epoch": 0.7675613934894346, "grad_norm": 21.370037078857422, "learning_rate": 6.968166224903749e-06, "loss": 0.5845, "step": 1008 }, { "epoch": 0.7683228631258329, "grad_norm": 27.036455154418945, "learning_rate": 6.962563592965996e-06, "loss": 0.8619, "step": 1009 }, { "epoch": 0.7690843327622311, "grad_norm": 28.398357391357422, "learning_rate": 6.956958046571564e-06, "loss": 0.8632, "step": 1010 }, { "epoch": 0.7698458023986293, "grad_norm": 31.176103591918945, "learning_rate": 6.951349594044839e-06, "loss": 0.7487, "step": 1011 }, { "epoch": 0.7706072720350277, "grad_norm": 24.909969329833984, "learning_rate": 6.945738243714508e-06, "loss": 0.6454, "step": 1012 }, { "epoch": 0.7713687416714259, "grad_norm": 28.9960994720459, "learning_rate": 6.940124003913576e-06, "loss": 0.7539, "step": 1013 }, { "epoch": 0.7721302113078241, "grad_norm": 24.397951126098633, "learning_rate": 6.934506882979324e-06, "loss": 0.7002, "step": 1014 }, { "epoch": 0.7728916809442223, "grad_norm": 29.717355728149414, "learning_rate": 6.928886889253321e-06, "loss": 0.8345, "step": 1015 }, { "epoch": 0.7736531505806206, "grad_norm": 29.544414520263672, "learning_rate": 6.923264031081405e-06, "loss": 0.8499, "step": 1016 }, { "epoch": 0.7744146202170189, "grad_norm": 23.689773559570312, "learning_rate": 6.917638316813657e-06, "loss": 0.6965, "step": 1017 }, { "epoch": 0.7751760898534171, "grad_norm": 30.964216232299805, "learning_rate": 6.912009754804408e-06, "loss": 0.7015, "step": 1018 }, { "epoch": 0.7759375594898154, "grad_norm": 27.801557540893555, "learning_rate": 6.906378353412214e-06, "loss": 0.5867, "step": 1019 }, { "epoch": 0.7766990291262136, "grad_norm": 31.580921173095703, "learning_rate": 6.900744120999852e-06, "loss": 0.7862, "step": 1020 }, { "epoch": 0.7774604987626118, "grad_norm": 33.233482360839844, "learning_rate": 6.895107065934294e-06, "loss": 0.8513, "step": 1021 }, { "epoch": 0.7782219683990101, "grad_norm": 23.308080673217773, "learning_rate": 6.889467196586717e-06, "loss": 0.7959, "step": 1022 }, { "epoch": 0.7789834380354084, "grad_norm": 30.126352310180664, "learning_rate": 6.8838245213324625e-06, "loss": 0.8789, "step": 1023 }, { "epoch": 0.7797449076718066, "grad_norm": 28.059972763061523, "learning_rate": 6.878179048551052e-06, "loss": 0.7098, "step": 1024 }, { "epoch": 0.7805063773082048, "grad_norm": 26.31889533996582, "learning_rate": 6.8725307866261525e-06, "loss": 0.6385, "step": 1025 }, { "epoch": 0.781267846944603, "grad_norm": 25.934246063232422, "learning_rate": 6.866879743945578e-06, "loss": 0.7846, "step": 1026 }, { "epoch": 0.7820293165810014, "grad_norm": 33.79861068725586, "learning_rate": 6.861225928901267e-06, "loss": 0.714, "step": 1027 }, { "epoch": 0.7827907862173996, "grad_norm": 24.4395751953125, "learning_rate": 6.855569349889282e-06, "loss": 0.6492, "step": 1028 }, { "epoch": 0.7835522558537978, "grad_norm": 33.73328399658203, "learning_rate": 6.849910015309784e-06, "loss": 0.8394, "step": 1029 }, { "epoch": 0.7843137254901961, "grad_norm": 22.482955932617188, "learning_rate": 6.84424793356703e-06, "loss": 0.6927, "step": 1030 }, { "epoch": 0.7850751951265943, "grad_norm": 33.160133361816406, "learning_rate": 6.838583113069353e-06, "loss": 0.764, "step": 1031 }, { "epoch": 0.7858366647629925, "grad_norm": 26.095256805419922, "learning_rate": 6.832915562229154e-06, "loss": 0.8866, "step": 1032 }, { "epoch": 0.7865981343993909, "grad_norm": 29.65294647216797, "learning_rate": 6.827245289462894e-06, "loss": 0.9335, "step": 1033 }, { "epoch": 0.7873596040357891, "grad_norm": 30.26475715637207, "learning_rate": 6.821572303191066e-06, "loss": 0.9639, "step": 1034 }, { "epoch": 0.7881210736721873, "grad_norm": 26.66080665588379, "learning_rate": 6.815896611838204e-06, "loss": 0.8112, "step": 1035 }, { "epoch": 0.7888825433085855, "grad_norm": 22.906566619873047, "learning_rate": 6.8102182238328496e-06, "loss": 0.6617, "step": 1036 }, { "epoch": 0.7896440129449838, "grad_norm": 26.10392189025879, "learning_rate": 6.804537147607556e-06, "loss": 0.8056, "step": 1037 }, { "epoch": 0.7904054825813821, "grad_norm": 25.522689819335938, "learning_rate": 6.798853391598861e-06, "loss": 0.5879, "step": 1038 }, { "epoch": 0.7911669522177803, "grad_norm": 45.19489288330078, "learning_rate": 6.793166964247291e-06, "loss": 0.8562, "step": 1039 }, { "epoch": 0.7919284218541786, "grad_norm": 25.769004821777344, "learning_rate": 6.7874778739973325e-06, "loss": 0.6726, "step": 1040 }, { "epoch": 0.7926898914905768, "grad_norm": 25.213132858276367, "learning_rate": 6.781786129297427e-06, "loss": 0.6337, "step": 1041 }, { "epoch": 0.793451361126975, "grad_norm": 31.456302642822266, "learning_rate": 6.776091738599963e-06, "loss": 0.7938, "step": 1042 }, { "epoch": 0.7942128307633733, "grad_norm": 37.52812194824219, "learning_rate": 6.7703947103612546e-06, "loss": 0.9799, "step": 1043 }, { "epoch": 0.7949743003997716, "grad_norm": 21.045913696289062, "learning_rate": 6.764695053041531e-06, "loss": 0.6326, "step": 1044 }, { "epoch": 0.7957357700361698, "grad_norm": 27.062198638916016, "learning_rate": 6.758992775104929e-06, "loss": 0.6531, "step": 1045 }, { "epoch": 0.796497239672568, "grad_norm": 32.12380599975586, "learning_rate": 6.753287885019475e-06, "loss": 0.8352, "step": 1046 }, { "epoch": 0.7972587093089663, "grad_norm": 26.72404670715332, "learning_rate": 6.747580391257077e-06, "loss": 0.5981, "step": 1047 }, { "epoch": 0.7980201789453646, "grad_norm": 25.282129287719727, "learning_rate": 6.741870302293507e-06, "loss": 0.7022, "step": 1048 }, { "epoch": 0.7987816485817628, "grad_norm": 27.5859432220459, "learning_rate": 6.736157626608391e-06, "loss": 1.0093, "step": 1049 }, { "epoch": 0.799543118218161, "grad_norm": 35.4617919921875, "learning_rate": 6.730442372685199e-06, "loss": 0.7315, "step": 1050 }, { "epoch": 0.8003045878545593, "grad_norm": 31.363588333129883, "learning_rate": 6.724724549011226e-06, "loss": 0.773, "step": 1051 }, { "epoch": 0.8010660574909575, "grad_norm": 22.282655715942383, "learning_rate": 6.719004164077586e-06, "loss": 0.6401, "step": 1052 }, { "epoch": 0.8018275271273558, "grad_norm": 25.954313278198242, "learning_rate": 6.7132812263791955e-06, "loss": 0.8356, "step": 1053 }, { "epoch": 0.8025889967637541, "grad_norm": 28.291763305664062, "learning_rate": 6.707555744414762e-06, "loss": 0.7005, "step": 1054 }, { "epoch": 0.8033504664001523, "grad_norm": 28.33548355102539, "learning_rate": 6.7018277266867725e-06, "loss": 0.6962, "step": 1055 }, { "epoch": 0.8041119360365505, "grad_norm": 23.257038116455078, "learning_rate": 6.696097181701476e-06, "loss": 0.6206, "step": 1056 }, { "epoch": 0.8048734056729487, "grad_norm": 21.727081298828125, "learning_rate": 6.690364117968879e-06, "loss": 0.6527, "step": 1057 }, { "epoch": 0.8056348753093471, "grad_norm": 26.588376998901367, "learning_rate": 6.684628544002727e-06, "loss": 0.7627, "step": 1058 }, { "epoch": 0.8063963449457453, "grad_norm": 32.86149597167969, "learning_rate": 6.6788904683204904e-06, "loss": 0.9152, "step": 1059 }, { "epoch": 0.8071578145821435, "grad_norm": 23.547754287719727, "learning_rate": 6.673149899443359e-06, "loss": 0.6943, "step": 1060 }, { "epoch": 0.8079192842185418, "grad_norm": 23.529922485351562, "learning_rate": 6.66740684589622e-06, "loss": 0.6628, "step": 1061 }, { "epoch": 0.80868075385494, "grad_norm": 25.855627059936523, "learning_rate": 6.6616613162076586e-06, "loss": 0.7292, "step": 1062 }, { "epoch": 0.8094422234913383, "grad_norm": 25.554996490478516, "learning_rate": 6.655913318909928e-06, "loss": 0.5966, "step": 1063 }, { "epoch": 0.8102036931277365, "grad_norm": 30.357669830322266, "learning_rate": 6.650162862538951e-06, "loss": 0.7575, "step": 1064 }, { "epoch": 0.8109651627641348, "grad_norm": 32.665164947509766, "learning_rate": 6.6444099556343e-06, "loss": 0.8165, "step": 1065 }, { "epoch": 0.811726632400533, "grad_norm": 25.377029418945312, "learning_rate": 6.63865460673919e-06, "loss": 0.6122, "step": 1066 }, { "epoch": 0.8124881020369312, "grad_norm": 21.833900451660156, "learning_rate": 6.632896824400458e-06, "loss": 0.635, "step": 1067 }, { "epoch": 0.8132495716733296, "grad_norm": 32.79814529418945, "learning_rate": 6.627136617168557e-06, "loss": 0.7899, "step": 1068 }, { "epoch": 0.8140110413097278, "grad_norm": 32.96546936035156, "learning_rate": 6.621373993597542e-06, "loss": 0.9737, "step": 1069 }, { "epoch": 0.814772510946126, "grad_norm": 26.516597747802734, "learning_rate": 6.615608962245053e-06, "loss": 0.8697, "step": 1070 }, { "epoch": 0.8155339805825242, "grad_norm": 29.84071159362793, "learning_rate": 6.609841531672309e-06, "loss": 0.805, "step": 1071 }, { "epoch": 0.8162954502189225, "grad_norm": 30.24439239501953, "learning_rate": 6.604071710444091e-06, "loss": 0.801, "step": 1072 }, { "epoch": 0.8170569198553208, "grad_norm": 27.35975456237793, "learning_rate": 6.5982995071287295e-06, "loss": 0.7307, "step": 1073 }, { "epoch": 0.817818389491719, "grad_norm": 22.919601440429688, "learning_rate": 6.592524930298092e-06, "loss": 0.6959, "step": 1074 }, { "epoch": 0.8185798591281173, "grad_norm": 22.750600814819336, "learning_rate": 6.586747988527571e-06, "loss": 0.6578, "step": 1075 }, { "epoch": 0.8193413287645155, "grad_norm": 25.95614242553711, "learning_rate": 6.5809686903960725e-06, "loss": 0.7552, "step": 1076 }, { "epoch": 0.8201027984009137, "grad_norm": 22.792787551879883, "learning_rate": 6.575187044486e-06, "loss": 0.6678, "step": 1077 }, { "epoch": 0.820864268037312, "grad_norm": 25.64997673034668, "learning_rate": 6.569403059383245e-06, "loss": 0.7129, "step": 1078 }, { "epoch": 0.8216257376737103, "grad_norm": 23.963544845581055, "learning_rate": 6.563616743677171e-06, "loss": 0.7109, "step": 1079 }, { "epoch": 0.8223872073101085, "grad_norm": 24.058998107910156, "learning_rate": 6.557828105960604e-06, "loss": 0.6249, "step": 1080 }, { "epoch": 0.8231486769465067, "grad_norm": 25.706268310546875, "learning_rate": 6.552037154829819e-06, "loss": 0.6757, "step": 1081 }, { "epoch": 0.823910146582905, "grad_norm": 25.184911727905273, "learning_rate": 6.546243898884523e-06, "loss": 0.6759, "step": 1082 }, { "epoch": 0.8246716162193033, "grad_norm": 28.541261672973633, "learning_rate": 6.540448346727849e-06, "loss": 0.8746, "step": 1083 }, { "epoch": 0.8254330858557015, "grad_norm": 24.3239803314209, "learning_rate": 6.534650506966338e-06, "loss": 0.6053, "step": 1084 }, { "epoch": 0.8261945554920997, "grad_norm": 23.462106704711914, "learning_rate": 6.52885038820993e-06, "loss": 0.8506, "step": 1085 }, { "epoch": 0.826956025128498, "grad_norm": 24.462833404541016, "learning_rate": 6.52304799907195e-06, "loss": 0.7742, "step": 1086 }, { "epoch": 0.8277174947648962, "grad_norm": 24.899484634399414, "learning_rate": 6.5172433481690866e-06, "loss": 0.8278, "step": 1087 }, { "epoch": 0.8284789644012945, "grad_norm": 26.425138473510742, "learning_rate": 6.5114364441214015e-06, "loss": 0.7248, "step": 1088 }, { "epoch": 0.8292404340376928, "grad_norm": 24.521818161010742, "learning_rate": 6.505627295552289e-06, "loss": 0.7338, "step": 1089 }, { "epoch": 0.830001903674091, "grad_norm": 28.257104873657227, "learning_rate": 6.499815911088486e-06, "loss": 0.8183, "step": 1090 }, { "epoch": 0.8307633733104892, "grad_norm": 28.80118751525879, "learning_rate": 6.4940022993600415e-06, "loss": 0.8351, "step": 1091 }, { "epoch": 0.8315248429468874, "grad_norm": 31.63120460510254, "learning_rate": 6.488186469000318e-06, "loss": 0.7681, "step": 1092 }, { "epoch": 0.8322863125832858, "grad_norm": 23.432279586791992, "learning_rate": 6.482368428645972e-06, "loss": 0.7409, "step": 1093 }, { "epoch": 0.833047782219684, "grad_norm": 26.677947998046875, "learning_rate": 6.476548186936939e-06, "loss": 0.7739, "step": 1094 }, { "epoch": 0.8338092518560822, "grad_norm": 23.054256439208984, "learning_rate": 6.470725752516428e-06, "loss": 0.655, "step": 1095 }, { "epoch": 0.8345707214924805, "grad_norm": 25.91546058654785, "learning_rate": 6.464901134030899e-06, "loss": 0.8314, "step": 1096 }, { "epoch": 0.8353321911288787, "grad_norm": 20.97500228881836, "learning_rate": 6.45907434013006e-06, "loss": 0.6261, "step": 1097 }, { "epoch": 0.836093660765277, "grad_norm": 20.611257553100586, "learning_rate": 6.453245379466845e-06, "loss": 0.6271, "step": 1098 }, { "epoch": 0.8368551304016753, "grad_norm": 29.445005416870117, "learning_rate": 6.447414260697412e-06, "loss": 0.74, "step": 1099 }, { "epoch": 0.8376166000380735, "grad_norm": 23.64499855041504, "learning_rate": 6.441580992481117e-06, "loss": 0.799, "step": 1100 }, { "epoch": 0.8383780696744717, "grad_norm": 27.02738380432129, "learning_rate": 6.435745583480514e-06, "loss": 0.6591, "step": 1101 }, { "epoch": 0.8391395393108699, "grad_norm": 30.891071319580078, "learning_rate": 6.429908042361329e-06, "loss": 0.5697, "step": 1102 }, { "epoch": 0.8399010089472683, "grad_norm": 22.450510025024414, "learning_rate": 6.424068377792465e-06, "loss": 0.5774, "step": 1103 }, { "epoch": 0.8406624785836665, "grad_norm": 32.449337005615234, "learning_rate": 6.4182265984459645e-06, "loss": 0.831, "step": 1104 }, { "epoch": 0.8414239482200647, "grad_norm": 26.812030792236328, "learning_rate": 6.412382712997024e-06, "loss": 0.7699, "step": 1105 }, { "epoch": 0.842185417856463, "grad_norm": 25.392494201660156, "learning_rate": 6.406536730123958e-06, "loss": 0.7385, "step": 1106 }, { "epoch": 0.8429468874928612, "grad_norm": 58.397918701171875, "learning_rate": 6.400688658508201e-06, "loss": 0.81, "step": 1107 }, { "epoch": 0.8437083571292595, "grad_norm": 22.630483627319336, "learning_rate": 6.394838506834286e-06, "loss": 0.604, "step": 1108 }, { "epoch": 0.8444698267656577, "grad_norm": 23.90361785888672, "learning_rate": 6.388986283789837e-06, "loss": 0.7137, "step": 1109 }, { "epoch": 0.845231296402056, "grad_norm": 23.948640823364258, "learning_rate": 6.383131998065555e-06, "loss": 0.5811, "step": 1110 }, { "epoch": 0.8459927660384542, "grad_norm": 26.288042068481445, "learning_rate": 6.3772756583551985e-06, "loss": 0.8181, "step": 1111 }, { "epoch": 0.8467542356748524, "grad_norm": 26.572824478149414, "learning_rate": 6.371417273355587e-06, "loss": 0.6547, "step": 1112 }, { "epoch": 0.8475157053112508, "grad_norm": 27.283893585205078, "learning_rate": 6.365556851766565e-06, "loss": 0.6623, "step": 1113 }, { "epoch": 0.848277174947649, "grad_norm": 21.808853149414062, "learning_rate": 6.35969440229101e-06, "loss": 0.6685, "step": 1114 }, { "epoch": 0.8490386445840472, "grad_norm": 33.87482452392578, "learning_rate": 6.353829933634806e-06, "loss": 0.8288, "step": 1115 }, { "epoch": 0.8498001142204454, "grad_norm": 29.986732482910156, "learning_rate": 6.3479634545068415e-06, "loss": 0.876, "step": 1116 }, { "epoch": 0.8505615838568437, "grad_norm": 26.11766242980957, "learning_rate": 6.342094973618982e-06, "loss": 0.73, "step": 1117 }, { "epoch": 0.851323053493242, "grad_norm": 22.53919219970703, "learning_rate": 6.3362244996860734e-06, "loss": 0.6986, "step": 1118 }, { "epoch": 0.8520845231296402, "grad_norm": 27.49944305419922, "learning_rate": 6.330352041425916e-06, "loss": 0.8696, "step": 1119 }, { "epoch": 0.8528459927660385, "grad_norm": 30.59848976135254, "learning_rate": 6.324477607559261e-06, "loss": 0.8021, "step": 1120 }, { "epoch": 0.8536074624024367, "grad_norm": 27.27004051208496, "learning_rate": 6.31860120680979e-06, "loss": 0.758, "step": 1121 }, { "epoch": 0.8543689320388349, "grad_norm": 33.175846099853516, "learning_rate": 6.3127228479041095e-06, "loss": 0.6717, "step": 1122 }, { "epoch": 0.8551304016752332, "grad_norm": 25.63560676574707, "learning_rate": 6.306842539571729e-06, "loss": 0.5751, "step": 1123 }, { "epoch": 0.8558918713116315, "grad_norm": 26.582700729370117, "learning_rate": 6.3009602905450564e-06, "loss": 0.7455, "step": 1124 }, { "epoch": 0.8566533409480297, "grad_norm": 23.46794319152832, "learning_rate": 6.295076109559381e-06, "loss": 0.7402, "step": 1125 }, { "epoch": 0.8574148105844279, "grad_norm": 22.42741584777832, "learning_rate": 6.2891900053528594e-06, "loss": 0.6493, "step": 1126 }, { "epoch": 0.8581762802208261, "grad_norm": 31.379404067993164, "learning_rate": 6.283301986666508e-06, "loss": 1.0314, "step": 1127 }, { "epoch": 0.8589377498572245, "grad_norm": 26.408973693847656, "learning_rate": 6.27741206224418e-06, "loss": 0.7654, "step": 1128 }, { "epoch": 0.8596992194936227, "grad_norm": 25.700510025024414, "learning_rate": 6.271520240832567e-06, "loss": 0.7395, "step": 1129 }, { "epoch": 0.8604606891300209, "grad_norm": 32.66632843017578, "learning_rate": 6.265626531181168e-06, "loss": 0.8508, "step": 1130 }, { "epoch": 0.8612221587664192, "grad_norm": 22.692630767822266, "learning_rate": 6.259730942042296e-06, "loss": 0.7049, "step": 1131 }, { "epoch": 0.8619836284028174, "grad_norm": 19.742332458496094, "learning_rate": 6.2538334821710476e-06, "loss": 0.6357, "step": 1132 }, { "epoch": 0.8627450980392157, "grad_norm": 24.254117965698242, "learning_rate": 6.2479341603253e-06, "loss": 0.6543, "step": 1133 }, { "epoch": 0.863506567675614, "grad_norm": 26.565750122070312, "learning_rate": 6.242032985265697e-06, "loss": 0.9263, "step": 1134 }, { "epoch": 0.8642680373120122, "grad_norm": 25.064411163330078, "learning_rate": 6.236129965755633e-06, "loss": 0.7509, "step": 1135 }, { "epoch": 0.8650295069484104, "grad_norm": 30.72630500793457, "learning_rate": 6.230225110561241e-06, "loss": 0.9994, "step": 1136 }, { "epoch": 0.8657909765848086, "grad_norm": 22.605377197265625, "learning_rate": 6.22431842845138e-06, "loss": 0.6749, "step": 1137 }, { "epoch": 0.866552446221207, "grad_norm": 24.304431915283203, "learning_rate": 6.218409928197624e-06, "loss": 0.5875, "step": 1138 }, { "epoch": 0.8673139158576052, "grad_norm": 27.662187576293945, "learning_rate": 6.212499618574245e-06, "loss": 0.7709, "step": 1139 }, { "epoch": 0.8680753854940034, "grad_norm": 30.374610900878906, "learning_rate": 6.206587508358202e-06, "loss": 0.8106, "step": 1140 }, { "epoch": 0.8688368551304017, "grad_norm": 25.036169052124023, "learning_rate": 6.20067360632913e-06, "loss": 0.5386, "step": 1141 }, { "epoch": 0.8695983247667999, "grad_norm": 24.271564483642578, "learning_rate": 6.194757921269323e-06, "loss": 0.6632, "step": 1142 }, { "epoch": 0.8703597944031982, "grad_norm": 26.905012130737305, "learning_rate": 6.18884046196372e-06, "loss": 0.9513, "step": 1143 }, { "epoch": 0.8711212640395964, "grad_norm": 30.236791610717773, "learning_rate": 6.182921237199903e-06, "loss": 0.9973, "step": 1144 }, { "epoch": 0.8718827336759947, "grad_norm": 25.394094467163086, "learning_rate": 6.1770002557680675e-06, "loss": 0.737, "step": 1145 }, { "epoch": 0.8726442033123929, "grad_norm": 24.748056411743164, "learning_rate": 6.171077526461022e-06, "loss": 0.6292, "step": 1146 }, { "epoch": 0.8734056729487911, "grad_norm": 24.48686408996582, "learning_rate": 6.165153058074172e-06, "loss": 0.5377, "step": 1147 }, { "epoch": 0.8741671425851895, "grad_norm": 24.33057403564453, "learning_rate": 6.159226859405498e-06, "loss": 0.72, "step": 1148 }, { "epoch": 0.8749286122215877, "grad_norm": 23.461620330810547, "learning_rate": 6.15329893925556e-06, "loss": 0.5776, "step": 1149 }, { "epoch": 0.8756900818579859, "grad_norm": 26.3385066986084, "learning_rate": 6.147369306427466e-06, "loss": 0.7486, "step": 1150 }, { "epoch": 0.8764515514943841, "grad_norm": 28.738296508789062, "learning_rate": 6.141437969726875e-06, "loss": 0.9114, "step": 1151 }, { "epoch": 0.8772130211307824, "grad_norm": 28.215402603149414, "learning_rate": 6.135504937961968e-06, "loss": 0.6128, "step": 1152 }, { "epoch": 0.8779744907671807, "grad_norm": 28.2916259765625, "learning_rate": 6.12957021994345e-06, "loss": 0.8354, "step": 1153 }, { "epoch": 0.8787359604035789, "grad_norm": 28.98050880432129, "learning_rate": 6.123633824484529e-06, "loss": 0.937, "step": 1154 }, { "epoch": 0.8794974300399772, "grad_norm": 23.90348243713379, "learning_rate": 6.1176957604009004e-06, "loss": 0.75, "step": 1155 }, { "epoch": 0.8802588996763754, "grad_norm": 23.041603088378906, "learning_rate": 6.1117560365107395e-06, "loss": 0.5505, "step": 1156 }, { "epoch": 0.8810203693127736, "grad_norm": 24.29511833190918, "learning_rate": 6.105814661634688e-06, "loss": 0.7505, "step": 1157 }, { "epoch": 0.881781838949172, "grad_norm": 33.03423309326172, "learning_rate": 6.0998716445958384e-06, "loss": 0.7692, "step": 1158 }, { "epoch": 0.8825433085855702, "grad_norm": 27.23056983947754, "learning_rate": 6.093926994219721e-06, "loss": 0.7644, "step": 1159 }, { "epoch": 0.8833047782219684, "grad_norm": 25.457029342651367, "learning_rate": 6.0879807193342934e-06, "loss": 0.6943, "step": 1160 }, { "epoch": 0.8840662478583666, "grad_norm": 29.422870635986328, "learning_rate": 6.082032828769923e-06, "loss": 0.7901, "step": 1161 }, { "epoch": 0.8848277174947649, "grad_norm": 28.38987159729004, "learning_rate": 6.076083331359378e-06, "loss": 0.6902, "step": 1162 }, { "epoch": 0.8855891871311632, "grad_norm": 26.796110153198242, "learning_rate": 6.0701322359378136e-06, "loss": 0.712, "step": 1163 }, { "epoch": 0.8863506567675614, "grad_norm": 26.46685218811035, "learning_rate": 6.064179551342757e-06, "loss": 0.7801, "step": 1164 }, { "epoch": 0.8871121264039596, "grad_norm": 27.6890926361084, "learning_rate": 6.058225286414097e-06, "loss": 0.7373, "step": 1165 }, { "epoch": 0.8878735960403579, "grad_norm": 30.19915771484375, "learning_rate": 6.052269449994066e-06, "loss": 0.8487, "step": 1166 }, { "epoch": 0.8886350656767561, "grad_norm": 29.162994384765625, "learning_rate": 6.046312050927233e-06, "loss": 0.715, "step": 1167 }, { "epoch": 0.8893965353131544, "grad_norm": 26.86615562438965, "learning_rate": 6.040353098060486e-06, "loss": 0.6446, "step": 1168 }, { "epoch": 0.8901580049495527, "grad_norm": 21.35901641845703, "learning_rate": 6.034392600243022e-06, "loss": 0.5702, "step": 1169 }, { "epoch": 0.8909194745859509, "grad_norm": 21.94304847717285, "learning_rate": 6.028430566326331e-06, "loss": 0.7641, "step": 1170 }, { "epoch": 0.8916809442223491, "grad_norm": 31.8968448638916, "learning_rate": 6.022467005164185e-06, "loss": 0.8868, "step": 1171 }, { "epoch": 0.8924424138587473, "grad_norm": 29.475032806396484, "learning_rate": 6.01650192561262e-06, "loss": 0.8715, "step": 1172 }, { "epoch": 0.8932038834951457, "grad_norm": 30.826934814453125, "learning_rate": 6.010535336529935e-06, "loss": 0.7661, "step": 1173 }, { "epoch": 0.8939653531315439, "grad_norm": 25.327468872070312, "learning_rate": 6.004567246776663e-06, "loss": 0.6466, "step": 1174 }, { "epoch": 0.8947268227679421, "grad_norm": 28.873245239257812, "learning_rate": 5.998597665215571e-06, "loss": 0.8052, "step": 1175 }, { "epoch": 0.8954882924043404, "grad_norm": 31.27683448791504, "learning_rate": 5.992626600711637e-06, "loss": 0.7915, "step": 1176 }, { "epoch": 0.8962497620407386, "grad_norm": 29.112852096557617, "learning_rate": 5.986654062132043e-06, "loss": 0.802, "step": 1177 }, { "epoch": 0.8970112316771369, "grad_norm": 22.414955139160156, "learning_rate": 5.980680058346161e-06, "loss": 0.568, "step": 1178 }, { "epoch": 0.8977727013135351, "grad_norm": 29.032291412353516, "learning_rate": 5.974704598225537e-06, "loss": 0.6135, "step": 1179 }, { "epoch": 0.8985341709499334, "grad_norm": 23.16351890563965, "learning_rate": 5.968727690643884e-06, "loss": 0.5811, "step": 1180 }, { "epoch": 0.8992956405863316, "grad_norm": 25.667726516723633, "learning_rate": 5.962749344477056e-06, "loss": 0.7175, "step": 1181 }, { "epoch": 0.9000571102227298, "grad_norm": 23.955896377563477, "learning_rate": 5.956769568603054e-06, "loss": 0.7257, "step": 1182 }, { "epoch": 0.9008185798591282, "grad_norm": 31.422351837158203, "learning_rate": 5.950788371901992e-06, "loss": 0.7054, "step": 1183 }, { "epoch": 0.9015800494955264, "grad_norm": 27.23788070678711, "learning_rate": 5.944805763256102e-06, "loss": 0.7344, "step": 1184 }, { "epoch": 0.9023415191319246, "grad_norm": 23.782405853271484, "learning_rate": 5.938821751549707e-06, "loss": 0.5489, "step": 1185 }, { "epoch": 0.9031029887683228, "grad_norm": 32.04588317871094, "learning_rate": 5.932836345669217e-06, "loss": 0.7024, "step": 1186 }, { "epoch": 0.9038644584047211, "grad_norm": 26.27791976928711, "learning_rate": 5.9268495545031136e-06, "loss": 0.7602, "step": 1187 }, { "epoch": 0.9046259280411194, "grad_norm": 28.401838302612305, "learning_rate": 5.920861386941931e-06, "loss": 0.8044, "step": 1188 }, { "epoch": 0.9053873976775176, "grad_norm": 31.61398696899414, "learning_rate": 5.914871851878249e-06, "loss": 0.7006, "step": 1189 }, { "epoch": 0.9061488673139159, "grad_norm": 34.011348724365234, "learning_rate": 5.9088809582066795e-06, "loss": 0.819, "step": 1190 }, { "epoch": 0.9069103369503141, "grad_norm": 30.257753372192383, "learning_rate": 5.902888714823856e-06, "loss": 0.8277, "step": 1191 }, { "epoch": 0.9076718065867123, "grad_norm": 31.04530906677246, "learning_rate": 5.896895130628404e-06, "loss": 0.7759, "step": 1192 }, { "epoch": 0.9084332762231107, "grad_norm": 40.253196716308594, "learning_rate": 5.890900214520955e-06, "loss": 0.7379, "step": 1193 }, { "epoch": 0.9091947458595089, "grad_norm": 28.528383255004883, "learning_rate": 5.884903975404105e-06, "loss": 0.881, "step": 1194 }, { "epoch": 0.9099562154959071, "grad_norm": 28.120880126953125, "learning_rate": 5.8789064221824276e-06, "loss": 0.8019, "step": 1195 }, { "epoch": 0.9107176851323053, "grad_norm": 27.11699676513672, "learning_rate": 5.872907563762435e-06, "loss": 0.6886, "step": 1196 }, { "epoch": 0.9114791547687036, "grad_norm": 30.131195068359375, "learning_rate": 5.86690740905259e-06, "loss": 1.0627, "step": 1197 }, { "epoch": 0.9122406244051019, "grad_norm": 24.216678619384766, "learning_rate": 5.8609059669632695e-06, "loss": 0.5997, "step": 1198 }, { "epoch": 0.9130020940415001, "grad_norm": 29.27970314025879, "learning_rate": 5.854903246406771e-06, "loss": 0.8601, "step": 1199 }, { "epoch": 0.9137635636778983, "grad_norm": 29.73092269897461, "learning_rate": 5.848899256297282e-06, "loss": 0.8644, "step": 1200 }, { "epoch": 0.9145250333142966, "grad_norm": 27.606903076171875, "learning_rate": 5.842894005550885e-06, "loss": 0.7899, "step": 1201 }, { "epoch": 0.9152865029506948, "grad_norm": 28.840871810913086, "learning_rate": 5.8368875030855275e-06, "loss": 0.8073, "step": 1202 }, { "epoch": 0.9160479725870931, "grad_norm": 39.332088470458984, "learning_rate": 5.830879757821015e-06, "loss": 0.8845, "step": 1203 }, { "epoch": 0.9168094422234914, "grad_norm": 30.630475997924805, "learning_rate": 5.824870778679008e-06, "loss": 0.6556, "step": 1204 }, { "epoch": 0.9175709118598896, "grad_norm": 24.6275634765625, "learning_rate": 5.8188605745829855e-06, "loss": 0.8334, "step": 1205 }, { "epoch": 0.9183323814962878, "grad_norm": 26.482088088989258, "learning_rate": 5.812849154458259e-06, "loss": 0.8319, "step": 1206 }, { "epoch": 0.919093851132686, "grad_norm": 21.814990997314453, "learning_rate": 5.806836527231935e-06, "loss": 0.6287, "step": 1207 }, { "epoch": 0.9198553207690844, "grad_norm": 26.506677627563477, "learning_rate": 5.800822701832921e-06, "loss": 0.797, "step": 1208 }, { "epoch": 0.9206167904054826, "grad_norm": 33.39175033569336, "learning_rate": 5.794807687191897e-06, "loss": 0.8004, "step": 1209 }, { "epoch": 0.9213782600418808, "grad_norm": 27.25116539001465, "learning_rate": 5.788791492241316e-06, "loss": 0.7772, "step": 1210 }, { "epoch": 0.9221397296782791, "grad_norm": 21.633432388305664, "learning_rate": 5.782774125915375e-06, "loss": 0.5232, "step": 1211 }, { "epoch": 0.9229011993146773, "grad_norm": 24.91562271118164, "learning_rate": 5.77675559715002e-06, "loss": 0.8105, "step": 1212 }, { "epoch": 0.9236626689510756, "grad_norm": 26.940095901489258, "learning_rate": 5.770735914882916e-06, "loss": 0.8398, "step": 1213 }, { "epoch": 0.9244241385874739, "grad_norm": 26.2950439453125, "learning_rate": 5.764715088053445e-06, "loss": 0.5688, "step": 1214 }, { "epoch": 0.9251856082238721, "grad_norm": 30.187541961669922, "learning_rate": 5.758693125602688e-06, "loss": 0.8419, "step": 1215 }, { "epoch": 0.9259470778602703, "grad_norm": 26.059171676635742, "learning_rate": 5.75267003647341e-06, "loss": 0.5121, "step": 1216 }, { "epoch": 0.9267085474966685, "grad_norm": 20.288402557373047, "learning_rate": 5.746645829610052e-06, "loss": 0.5588, "step": 1217 }, { "epoch": 0.9274700171330669, "grad_norm": 26.386737823486328, "learning_rate": 5.740620513958714e-06, "loss": 0.8128, "step": 1218 }, { "epoch": 0.9282314867694651, "grad_norm": 22.0696964263916, "learning_rate": 5.734594098467144e-06, "loss": 0.718, "step": 1219 }, { "epoch": 0.9289929564058633, "grad_norm": 21.758426666259766, "learning_rate": 5.728566592084718e-06, "loss": 0.5301, "step": 1220 }, { "epoch": 0.9297544260422616, "grad_norm": 26.532880783081055, "learning_rate": 5.722538003762439e-06, "loss": 0.698, "step": 1221 }, { "epoch": 0.9305158956786598, "grad_norm": 22.46722984313965, "learning_rate": 5.716508342452911e-06, "loss": 0.628, "step": 1222 }, { "epoch": 0.9312773653150581, "grad_norm": 24.560373306274414, "learning_rate": 5.710477617110337e-06, "loss": 0.6406, "step": 1223 }, { "epoch": 0.9320388349514563, "grad_norm": 27.06344985961914, "learning_rate": 5.7044458366904934e-06, "loss": 0.7911, "step": 1224 }, { "epoch": 0.9328003045878546, "grad_norm": 32.2224235534668, "learning_rate": 5.698413010150729e-06, "loss": 0.8885, "step": 1225 }, { "epoch": 0.9335617742242528, "grad_norm": 25.872486114501953, "learning_rate": 5.692379146449944e-06, "loss": 0.6416, "step": 1226 }, { "epoch": 0.934323243860651, "grad_norm": 35.66846466064453, "learning_rate": 5.686344254548579e-06, "loss": 0.9318, "step": 1227 }, { "epoch": 0.9350847134970494, "grad_norm": 28.10585594177246, "learning_rate": 5.6803083434086e-06, "loss": 0.7696, "step": 1228 }, { "epoch": 0.9358461831334476, "grad_norm": 30.744041442871094, "learning_rate": 5.674271421993491e-06, "loss": 0.5777, "step": 1229 }, { "epoch": 0.9366076527698458, "grad_norm": 21.80347442626953, "learning_rate": 5.668233499268231e-06, "loss": 0.6889, "step": 1230 }, { "epoch": 0.937369122406244, "grad_norm": 31.691043853759766, "learning_rate": 5.662194584199288e-06, "loss": 0.7792, "step": 1231 }, { "epoch": 0.9381305920426423, "grad_norm": 28.97787857055664, "learning_rate": 5.656154685754605e-06, "loss": 0.84, "step": 1232 }, { "epoch": 0.9388920616790406, "grad_norm": 21.709415435791016, "learning_rate": 5.6501138129035834e-06, "loss": 0.6616, "step": 1233 }, { "epoch": 0.9396535313154388, "grad_norm": 23.397506713867188, "learning_rate": 5.644071974617075e-06, "loss": 0.7269, "step": 1234 }, { "epoch": 0.940415000951837, "grad_norm": 38.36065673828125, "learning_rate": 5.638029179867359e-06, "loss": 0.9993, "step": 1235 }, { "epoch": 0.9411764705882353, "grad_norm": 26.851713180541992, "learning_rate": 5.63198543762814e-06, "loss": 0.8253, "step": 1236 }, { "epoch": 0.9419379402246335, "grad_norm": 32.275081634521484, "learning_rate": 5.625940756874531e-06, "loss": 0.9174, "step": 1237 }, { "epoch": 0.9426994098610318, "grad_norm": 30.562349319458008, "learning_rate": 5.619895146583032e-06, "loss": 0.8554, "step": 1238 }, { "epoch": 0.9434608794974301, "grad_norm": 23.26239013671875, "learning_rate": 5.613848615731531e-06, "loss": 0.7101, "step": 1239 }, { "epoch": 0.9442223491338283, "grad_norm": 23.173625946044922, "learning_rate": 5.607801173299278e-06, "loss": 0.7692, "step": 1240 }, { "epoch": 0.9449838187702265, "grad_norm": 24.91485595703125, "learning_rate": 5.601752828266878e-06, "loss": 0.8074, "step": 1241 }, { "epoch": 0.9457452884066248, "grad_norm": 32.00895690917969, "learning_rate": 5.595703589616277e-06, "loss": 0.863, "step": 1242 }, { "epoch": 0.9465067580430231, "grad_norm": 28.49915313720703, "learning_rate": 5.589653466330748e-06, "loss": 0.8436, "step": 1243 }, { "epoch": 0.9472682276794213, "grad_norm": 29.013669967651367, "learning_rate": 5.583602467394877e-06, "loss": 0.7377, "step": 1244 }, { "epoch": 0.9480296973158195, "grad_norm": 22.557308197021484, "learning_rate": 5.577550601794551e-06, "loss": 0.6627, "step": 1245 }, { "epoch": 0.9487911669522178, "grad_norm": 23.932159423828125, "learning_rate": 5.5714978785169434e-06, "loss": 0.9372, "step": 1246 }, { "epoch": 0.949552636588616, "grad_norm": 21.610549926757812, "learning_rate": 5.565444306550503e-06, "loss": 0.7085, "step": 1247 }, { "epoch": 0.9503141062250143, "grad_norm": 25.1000919342041, "learning_rate": 5.5593898948849365e-06, "loss": 0.7516, "step": 1248 }, { "epoch": 0.9510755758614126, "grad_norm": 27.611072540283203, "learning_rate": 5.5533346525111985e-06, "loss": 0.9046, "step": 1249 }, { "epoch": 0.9518370454978108, "grad_norm": 29.340579986572266, "learning_rate": 5.547278588421478e-06, "loss": 0.6753, "step": 1250 }, { "epoch": 0.952598515134209, "grad_norm": 29.948381423950195, "learning_rate": 5.541221711609183e-06, "loss": 0.7346, "step": 1251 }, { "epoch": 0.9533599847706072, "grad_norm": 24.31911849975586, "learning_rate": 5.5351640310689295e-06, "loss": 0.6716, "step": 1252 }, { "epoch": 0.9541214544070056, "grad_norm": 23.80052947998047, "learning_rate": 5.529105555796527e-06, "loss": 0.6168, "step": 1253 }, { "epoch": 0.9548829240434038, "grad_norm": 31.234840393066406, "learning_rate": 5.523046294788963e-06, "loss": 0.9446, "step": 1254 }, { "epoch": 0.955644393679802, "grad_norm": 22.28814125061035, "learning_rate": 5.516986257044397e-06, "loss": 0.7277, "step": 1255 }, { "epoch": 0.9564058633162003, "grad_norm": 23.948471069335938, "learning_rate": 5.510925451562134e-06, "loss": 0.5267, "step": 1256 }, { "epoch": 0.9571673329525985, "grad_norm": 25.476354598999023, "learning_rate": 5.504863887342629e-06, "loss": 0.6277, "step": 1257 }, { "epoch": 0.9579288025889967, "grad_norm": 22.085304260253906, "learning_rate": 5.498801573387454e-06, "loss": 0.6413, "step": 1258 }, { "epoch": 0.958690272225395, "grad_norm": 27.318944931030273, "learning_rate": 5.492738518699303e-06, "loss": 0.6536, "step": 1259 }, { "epoch": 0.9594517418617933, "grad_norm": 33.47896957397461, "learning_rate": 5.486674732281963e-06, "loss": 0.9644, "step": 1260 }, { "epoch": 0.9602132114981915, "grad_norm": 25.332130432128906, "learning_rate": 5.48061022314031e-06, "loss": 0.6966, "step": 1261 }, { "epoch": 0.9609746811345897, "grad_norm": 29.443384170532227, "learning_rate": 5.4745450002802955e-06, "loss": 0.7906, "step": 1262 }, { "epoch": 0.961736150770988, "grad_norm": 26.96221923828125, "learning_rate": 5.468479072708929e-06, "loss": 0.7849, "step": 1263 }, { "epoch": 0.9624976204073863, "grad_norm": 26.056095123291016, "learning_rate": 5.4624124494342655e-06, "loss": 0.5675, "step": 1264 }, { "epoch": 0.9632590900437845, "grad_norm": 28.76567840576172, "learning_rate": 5.456345139465396e-06, "loss": 0.6394, "step": 1265 }, { "epoch": 0.9640205596801827, "grad_norm": 26.48472023010254, "learning_rate": 5.450277151812427e-06, "loss": 0.7337, "step": 1266 }, { "epoch": 0.964782029316581, "grad_norm": 23.28859519958496, "learning_rate": 5.4442084954864785e-06, "loss": 0.7214, "step": 1267 }, { "epoch": 0.9655434989529792, "grad_norm": 27.207473754882812, "learning_rate": 5.438139179499654e-06, "loss": 0.6439, "step": 1268 }, { "epoch": 0.9663049685893775, "grad_norm": 25.156911849975586, "learning_rate": 5.432069212865044e-06, "loss": 0.719, "step": 1269 }, { "epoch": 0.9670664382257758, "grad_norm": 26.39915657043457, "learning_rate": 5.425998604596703e-06, "loss": 0.7213, "step": 1270 }, { "epoch": 0.967827907862174, "grad_norm": 28.437196731567383, "learning_rate": 5.419927363709636e-06, "loss": 0.8741, "step": 1271 }, { "epoch": 0.9685893774985722, "grad_norm": 26.70257568359375, "learning_rate": 5.413855499219794e-06, "loss": 0.7383, "step": 1272 }, { "epoch": 0.9693508471349704, "grad_norm": 22.76276969909668, "learning_rate": 5.407783020144044e-06, "loss": 0.5933, "step": 1273 }, { "epoch": 0.9701123167713688, "grad_norm": 23.470422744750977, "learning_rate": 5.401709935500176e-06, "loss": 0.629, "step": 1274 }, { "epoch": 0.970873786407767, "grad_norm": 22.049306869506836, "learning_rate": 5.395636254306871e-06, "loss": 0.6533, "step": 1275 }, { "epoch": 0.9716352560441652, "grad_norm": 29.921438217163086, "learning_rate": 5.389561985583704e-06, "loss": 0.585, "step": 1276 }, { "epoch": 0.9723967256805635, "grad_norm": 30.144283294677734, "learning_rate": 5.383487138351114e-06, "loss": 0.7616, "step": 1277 }, { "epoch": 0.9731581953169617, "grad_norm": 23.44175148010254, "learning_rate": 5.3774117216304034e-06, "loss": 0.701, "step": 1278 }, { "epoch": 0.97391966495336, "grad_norm": 33.73345184326172, "learning_rate": 5.371335744443722e-06, "loss": 0.9089, "step": 1279 }, { "epoch": 0.9746811345897582, "grad_norm": 32.97321319580078, "learning_rate": 5.365259215814048e-06, "loss": 0.8725, "step": 1280 }, { "epoch": 0.9754426042261565, "grad_norm": 29.173730850219727, "learning_rate": 5.3591821447651815e-06, "loss": 0.7169, "step": 1281 }, { "epoch": 0.9762040738625547, "grad_norm": 30.256772994995117, "learning_rate": 5.353104540321726e-06, "loss": 0.5812, "step": 1282 }, { "epoch": 0.9769655434989529, "grad_norm": 33.78550720214844, "learning_rate": 5.347026411509082e-06, "loss": 0.6862, "step": 1283 }, { "epoch": 0.9777270131353513, "grad_norm": 26.264739990234375, "learning_rate": 5.340947767353418e-06, "loss": 0.7688, "step": 1284 }, { "epoch": 0.9784884827717495, "grad_norm": 27.745128631591797, "learning_rate": 5.334868616881682e-06, "loss": 0.7675, "step": 1285 }, { "epoch": 0.9792499524081477, "grad_norm": 26.023317337036133, "learning_rate": 5.32878896912156e-06, "loss": 0.7348, "step": 1286 }, { "epoch": 0.9800114220445459, "grad_norm": 19.755701065063477, "learning_rate": 5.322708833101489e-06, "loss": 0.5437, "step": 1287 }, { "epoch": 0.9807728916809442, "grad_norm": 23.624074935913086, "learning_rate": 5.316628217850619e-06, "loss": 0.4966, "step": 1288 }, { "epoch": 0.9815343613173425, "grad_norm": 34.3329963684082, "learning_rate": 5.3105471323988234e-06, "loss": 0.905, "step": 1289 }, { "epoch": 0.9822958309537407, "grad_norm": 23.290987014770508, "learning_rate": 5.304465585776662e-06, "loss": 0.6532, "step": 1290 }, { "epoch": 0.983057300590139, "grad_norm": 24.961231231689453, "learning_rate": 5.298383587015393e-06, "loss": 0.7432, "step": 1291 }, { "epoch": 0.9838187702265372, "grad_norm": 28.71943473815918, "learning_rate": 5.292301145146933e-06, "loss": 0.7306, "step": 1292 }, { "epoch": 0.9845802398629354, "grad_norm": 16.87373924255371, "learning_rate": 5.286218269203867e-06, "loss": 0.5724, "step": 1293 }, { "epoch": 0.9853417094993338, "grad_norm": 31.654150009155273, "learning_rate": 5.280134968219415e-06, "loss": 0.9882, "step": 1294 }, { "epoch": 0.986103179135732, "grad_norm": 30.381277084350586, "learning_rate": 5.274051251227435e-06, "loss": 0.7664, "step": 1295 }, { "epoch": 0.9868646487721302, "grad_norm": 28.652997970581055, "learning_rate": 5.267967127262402e-06, "loss": 0.799, "step": 1296 }, { "epoch": 0.9876261184085284, "grad_norm": 28.27056312561035, "learning_rate": 5.261882605359392e-06, "loss": 0.889, "step": 1297 }, { "epoch": 0.9883875880449267, "grad_norm": 26.610063552856445, "learning_rate": 5.255797694554079e-06, "loss": 0.8926, "step": 1298 }, { "epoch": 0.989149057681325, "grad_norm": 26.535287857055664, "learning_rate": 5.249712403882704e-06, "loss": 0.8104, "step": 1299 }, { "epoch": 0.9899105273177232, "grad_norm": 19.606740951538086, "learning_rate": 5.243626742382082e-06, "loss": 0.6873, "step": 1300 }, { "epoch": 0.9906719969541214, "grad_norm": 23.807472229003906, "learning_rate": 5.2375407190895715e-06, "loss": 0.6477, "step": 1301 }, { "epoch": 0.9914334665905197, "grad_norm": 27.291015625, "learning_rate": 5.231454343043075e-06, "loss": 0.8188, "step": 1302 }, { "epoch": 0.9921949362269179, "grad_norm": 20.577680587768555, "learning_rate": 5.22536762328101e-06, "loss": 0.6778, "step": 1303 }, { "epoch": 0.9929564058633162, "grad_norm": 23.850460052490234, "learning_rate": 5.219280568842317e-06, "loss": 0.6774, "step": 1304 }, { "epoch": 0.9937178754997145, "grad_norm": 22.797754287719727, "learning_rate": 5.213193188766418e-06, "loss": 0.6328, "step": 1305 }, { "epoch": 0.9944793451361127, "grad_norm": 22.607013702392578, "learning_rate": 5.207105492093232e-06, "loss": 0.8442, "step": 1306 }, { "epoch": 0.9952408147725109, "grad_norm": 25.141313552856445, "learning_rate": 5.20101748786314e-06, "loss": 0.7963, "step": 1307 }, { "epoch": 0.9960022844089091, "grad_norm": 23.462364196777344, "learning_rate": 5.194929185116986e-06, "loss": 0.6818, "step": 1308 }, { "epoch": 0.9967637540453075, "grad_norm": 20.63343620300293, "learning_rate": 5.188840592896051e-06, "loss": 0.6168, "step": 1309 }, { "epoch": 0.9975252236817057, "grad_norm": 24.033830642700195, "learning_rate": 5.182751720242049e-06, "loss": 0.4904, "step": 1310 }, { "epoch": 0.9982866933181039, "grad_norm": 19.662151336669922, "learning_rate": 5.176662576197111e-06, "loss": 0.5352, "step": 1311 }, { "epoch": 0.9990481629545022, "grad_norm": 23.323802947998047, "learning_rate": 5.1705731698037695e-06, "loss": 0.8118, "step": 1312 }, { "epoch": 0.9998096325909004, "grad_norm": 26.61143684387207, "learning_rate": 5.16448351010495e-06, "loss": 0.7524, "step": 1313 }, { "epoch": 1.0, "grad_norm": 46.33402633666992, "learning_rate": 5.158393606143945e-06, "loss": 0.6167, "step": 1314 }, { "epoch": 1.0007614696363982, "grad_norm": 20.729055404663086, "learning_rate": 5.152303466964425e-06, "loss": 0.621, "step": 1315 }, { "epoch": 1.0015229392727965, "grad_norm": 24.62146759033203, "learning_rate": 5.146213101610393e-06, "loss": 0.6521, "step": 1316 }, { "epoch": 1.0022844089091947, "grad_norm": 20.44754981994629, "learning_rate": 5.140122519126201e-06, "loss": 0.6287, "step": 1317 }, { "epoch": 1.003045878545593, "grad_norm": 21.060617446899414, "learning_rate": 5.134031728556518e-06, "loss": 0.5237, "step": 1318 }, { "epoch": 1.0038073481819911, "grad_norm": 32.90572738647461, "learning_rate": 5.1279407389463194e-06, "loss": 0.9762, "step": 1319 }, { "epoch": 1.0045688178183896, "grad_norm": 24.62529945373535, "learning_rate": 5.121849559340882e-06, "loss": 0.6479, "step": 1320 }, { "epoch": 1.0053302874547878, "grad_norm": 21.510913848876953, "learning_rate": 5.115758198785761e-06, "loss": 0.5968, "step": 1321 }, { "epoch": 1.006091757091186, "grad_norm": 26.872161865234375, "learning_rate": 5.109666666326781e-06, "loss": 0.6928, "step": 1322 }, { "epoch": 1.0068532267275843, "grad_norm": 30.045848846435547, "learning_rate": 5.10357497101002e-06, "loss": 0.7937, "step": 1323 }, { "epoch": 1.0076146963639825, "grad_norm": 23.325647354125977, "learning_rate": 5.0974831218818045e-06, "loss": 0.6678, "step": 1324 }, { "epoch": 1.0083761660003807, "grad_norm": 23.447351455688477, "learning_rate": 5.09139112798868e-06, "loss": 0.5993, "step": 1325 }, { "epoch": 1.009137635636779, "grad_norm": 24.03450584411621, "learning_rate": 5.085298998377417e-06, "loss": 0.6243, "step": 1326 }, { "epoch": 1.0098991052731772, "grad_norm": 24.32756996154785, "learning_rate": 5.0792067420949775e-06, "loss": 0.6651, "step": 1327 }, { "epoch": 1.0106605749095754, "grad_norm": 27.687728881835938, "learning_rate": 5.073114368188521e-06, "loss": 0.6628, "step": 1328 }, { "epoch": 1.0114220445459736, "grad_norm": 25.241317749023438, "learning_rate": 5.067021885705372e-06, "loss": 0.7304, "step": 1329 }, { "epoch": 1.012183514182372, "grad_norm": 26.63682746887207, "learning_rate": 5.060929303693025e-06, "loss": 0.6242, "step": 1330 }, { "epoch": 1.0129449838187703, "grad_norm": 26.13730239868164, "learning_rate": 5.054836631199117e-06, "loss": 0.5472, "step": 1331 }, { "epoch": 1.0137064534551685, "grad_norm": 28.472431182861328, "learning_rate": 5.0487438772714235e-06, "loss": 0.8068, "step": 1332 }, { "epoch": 1.0144679230915667, "grad_norm": 28.824962615966797, "learning_rate": 5.042651050957836e-06, "loss": 0.7024, "step": 1333 }, { "epoch": 1.015229392727965, "grad_norm": 26.180198669433594, "learning_rate": 5.036558161306358e-06, "loss": 0.5555, "step": 1334 }, { "epoch": 1.0159908623643632, "grad_norm": 24.1535701751709, "learning_rate": 5.030465217365081e-06, "loss": 0.7168, "step": 1335 }, { "epoch": 1.0167523320007614, "grad_norm": 25.247085571289062, "learning_rate": 5.024372228182187e-06, "loss": 0.659, "step": 1336 }, { "epoch": 1.0175138016371597, "grad_norm": 36.04473114013672, "learning_rate": 5.018279202805916e-06, "loss": 0.8104, "step": 1337 }, { "epoch": 1.0182752712735579, "grad_norm": 23.826383590698242, "learning_rate": 5.012186150284565e-06, "loss": 0.5998, "step": 1338 }, { "epoch": 1.019036740909956, "grad_norm": 23.549758911132812, "learning_rate": 5.006093079666472e-06, "loss": 0.5823, "step": 1339 }, { "epoch": 1.0197982105463546, "grad_norm": 29.11300277709961, "learning_rate": 5e-06, "loss": 0.8563, "step": 1340 }, { "epoch": 1.0205596801827528, "grad_norm": 24.819345474243164, "learning_rate": 4.99390692033353e-06, "loss": 0.7238, "step": 1341 }, { "epoch": 1.021321149819151, "grad_norm": 22.405927658081055, "learning_rate": 4.987813849715436e-06, "loss": 0.638, "step": 1342 }, { "epoch": 1.0220826194555492, "grad_norm": 32.38371276855469, "learning_rate": 4.981720797194086e-06, "loss": 0.7509, "step": 1343 }, { "epoch": 1.0228440890919475, "grad_norm": 21.343875885009766, "learning_rate": 4.975627771817815e-06, "loss": 0.6095, "step": 1344 }, { "epoch": 1.0236055587283457, "grad_norm": 31.141231536865234, "learning_rate": 4.9695347826349204e-06, "loss": 0.7432, "step": 1345 }, { "epoch": 1.024367028364744, "grad_norm": 26.032634735107422, "learning_rate": 4.963441838693643e-06, "loss": 0.6004, "step": 1346 }, { "epoch": 1.0251284980011421, "grad_norm": 37.2136116027832, "learning_rate": 4.957348949042166e-06, "loss": 1.0375, "step": 1347 }, { "epoch": 1.0258899676375404, "grad_norm": 25.659366607666016, "learning_rate": 4.951256122728578e-06, "loss": 0.7755, "step": 1348 }, { "epoch": 1.0266514372739386, "grad_norm": 28.66021728515625, "learning_rate": 4.945163368800885e-06, "loss": 0.6546, "step": 1349 }, { "epoch": 1.027412906910337, "grad_norm": 26.520570755004883, "learning_rate": 4.939070696306976e-06, "loss": 0.6481, "step": 1350 }, { "epoch": 1.0281743765467353, "grad_norm": 23.912429809570312, "learning_rate": 4.9329781142946295e-06, "loss": 0.4242, "step": 1351 }, { "epoch": 1.0289358461831335, "grad_norm": 23.78453826904297, "learning_rate": 4.9268856318114816e-06, "loss": 0.5905, "step": 1352 }, { "epoch": 1.0296973158195317, "grad_norm": 27.954885482788086, "learning_rate": 4.9207932579050225e-06, "loss": 0.7001, "step": 1353 }, { "epoch": 1.03045878545593, "grad_norm": 27.79574203491211, "learning_rate": 4.914701001622585e-06, "loss": 0.7546, "step": 1354 }, { "epoch": 1.0312202550923282, "grad_norm": 27.99153709411621, "learning_rate": 4.9086088720113205e-06, "loss": 0.6991, "step": 1355 }, { "epoch": 1.0319817247287264, "grad_norm": 36.836402893066406, "learning_rate": 4.902516878118197e-06, "loss": 0.6438, "step": 1356 }, { "epoch": 1.0327431943651246, "grad_norm": 35.888099670410156, "learning_rate": 4.89642502898998e-06, "loss": 0.8721, "step": 1357 }, { "epoch": 1.0335046640015229, "grad_norm": 25.70697784423828, "learning_rate": 4.890333333673221e-06, "loss": 0.5341, "step": 1358 }, { "epoch": 1.034266133637921, "grad_norm": 28.39304542541504, "learning_rate": 4.884241801214241e-06, "loss": 0.706, "step": 1359 }, { "epoch": 1.0350276032743195, "grad_norm": 24.945363998413086, "learning_rate": 4.878150440659119e-06, "loss": 0.6264, "step": 1360 }, { "epoch": 1.0357890729107178, "grad_norm": 23.87784767150879, "learning_rate": 4.8720592610536805e-06, "loss": 0.6742, "step": 1361 }, { "epoch": 1.036550542547116, "grad_norm": 27.262670516967773, "learning_rate": 4.865968271443484e-06, "loss": 0.6899, "step": 1362 }, { "epoch": 1.0373120121835142, "grad_norm": 22.597061157226562, "learning_rate": 4.8598774808737995e-06, "loss": 0.572, "step": 1363 }, { "epoch": 1.0380734818199124, "grad_norm": 30.12322998046875, "learning_rate": 4.853786898389606e-06, "loss": 0.7398, "step": 1364 }, { "epoch": 1.0388349514563107, "grad_norm": 26.185476303100586, "learning_rate": 4.8476965330355764e-06, "loss": 0.7457, "step": 1365 }, { "epoch": 1.0395964210927089, "grad_norm": 25.132925033569336, "learning_rate": 4.841606393856056e-06, "loss": 0.7046, "step": 1366 }, { "epoch": 1.0403578907291071, "grad_norm": 26.63749122619629, "learning_rate": 4.835516489895053e-06, "loss": 0.5324, "step": 1367 }, { "epoch": 1.0411193603655053, "grad_norm": 28.17856788635254, "learning_rate": 4.829426830196231e-06, "loss": 0.7334, "step": 1368 }, { "epoch": 1.0418808300019036, "grad_norm": 26.967134475708008, "learning_rate": 4.82333742380289e-06, "loss": 0.7494, "step": 1369 }, { "epoch": 1.042642299638302, "grad_norm": 35.60799789428711, "learning_rate": 4.817248279757953e-06, "loss": 0.7949, "step": 1370 }, { "epoch": 1.0434037692747002, "grad_norm": 22.97283363342285, "learning_rate": 4.811159407103952e-06, "loss": 0.616, "step": 1371 }, { "epoch": 1.0441652389110985, "grad_norm": 35.12429428100586, "learning_rate": 4.805070814883015e-06, "loss": 0.819, "step": 1372 }, { "epoch": 1.0449267085474967, "grad_norm": 31.06621742248535, "learning_rate": 4.798982512136861e-06, "loss": 0.722, "step": 1373 }, { "epoch": 1.045688178183895, "grad_norm": 25.281723022460938, "learning_rate": 4.79289450790677e-06, "loss": 0.6381, "step": 1374 }, { "epoch": 1.0464496478202931, "grad_norm": 24.8349552154541, "learning_rate": 4.786806811233583e-06, "loss": 0.7381, "step": 1375 }, { "epoch": 1.0472111174566914, "grad_norm": 29.945003509521484, "learning_rate": 4.780719431157685e-06, "loss": 0.7243, "step": 1376 }, { "epoch": 1.0479725870930896, "grad_norm": 30.135971069335938, "learning_rate": 4.774632376718991e-06, "loss": 0.7976, "step": 1377 }, { "epoch": 1.0487340567294878, "grad_norm": 24.988666534423828, "learning_rate": 4.768545656956928e-06, "loss": 0.5811, "step": 1378 }, { "epoch": 1.049495526365886, "grad_norm": 18.137149810791016, "learning_rate": 4.7624592809104285e-06, "loss": 0.5986, "step": 1379 }, { "epoch": 1.0502569960022845, "grad_norm": 24.05771827697754, "learning_rate": 4.75637325761792e-06, "loss": 0.5975, "step": 1380 }, { "epoch": 1.0510184656386827, "grad_norm": 33.43357467651367, "learning_rate": 4.7502875961172975e-06, "loss": 0.9223, "step": 1381 }, { "epoch": 1.051779935275081, "grad_norm": 31.144792556762695, "learning_rate": 4.744202305445924e-06, "loss": 0.9286, "step": 1382 }, { "epoch": 1.0525414049114792, "grad_norm": 33.884578704833984, "learning_rate": 4.738117394640607e-06, "loss": 0.7424, "step": 1383 }, { "epoch": 1.0533028745478774, "grad_norm": 28.518062591552734, "learning_rate": 4.7320328727376e-06, "loss": 0.7627, "step": 1384 }, { "epoch": 1.0540643441842756, "grad_norm": 24.260276794433594, "learning_rate": 4.725948748772568e-06, "loss": 0.724, "step": 1385 }, { "epoch": 1.0548258138206739, "grad_norm": 23.066709518432617, "learning_rate": 4.7198650317805885e-06, "loss": 0.6625, "step": 1386 }, { "epoch": 1.055587283457072, "grad_norm": 25.143184661865234, "learning_rate": 4.713781730796136e-06, "loss": 0.6189, "step": 1387 }, { "epoch": 1.0563487530934703, "grad_norm": 21.390766143798828, "learning_rate": 4.707698854853068e-06, "loss": 0.6507, "step": 1388 }, { "epoch": 1.0571102227298685, "grad_norm": 29.9796142578125, "learning_rate": 4.701616412984609e-06, "loss": 0.7892, "step": 1389 }, { "epoch": 1.057871692366267, "grad_norm": 24.71466636657715, "learning_rate": 4.695534414223337e-06, "loss": 0.6041, "step": 1390 }, { "epoch": 1.0586331620026652, "grad_norm": 28.7713680267334, "learning_rate": 4.689452867601178e-06, "loss": 0.7806, "step": 1391 }, { "epoch": 1.0593946316390634, "grad_norm": 34.792259216308594, "learning_rate": 4.683371782149382e-06, "loss": 0.6914, "step": 1392 }, { "epoch": 1.0601561012754617, "grad_norm": 29.722267150878906, "learning_rate": 4.6772911668985135e-06, "loss": 0.6494, "step": 1393 }, { "epoch": 1.06091757091186, "grad_norm": 21.27020263671875, "learning_rate": 4.67121103087844e-06, "loss": 0.4866, "step": 1394 }, { "epoch": 1.0616790405482581, "grad_norm": 25.263330459594727, "learning_rate": 4.665131383118319e-06, "loss": 0.6472, "step": 1395 }, { "epoch": 1.0624405101846564, "grad_norm": 23.385417938232422, "learning_rate": 4.659052232646583e-06, "loss": 0.5967, "step": 1396 }, { "epoch": 1.0632019798210546, "grad_norm": 25.676136016845703, "learning_rate": 4.652973588490921e-06, "loss": 0.6093, "step": 1397 }, { "epoch": 1.0639634494574528, "grad_norm": 24.565771102905273, "learning_rate": 4.6468954596782736e-06, "loss": 0.7276, "step": 1398 }, { "epoch": 1.064724919093851, "grad_norm": 33.12934494018555, "learning_rate": 4.640817855234819e-06, "loss": 0.6635, "step": 1399 }, { "epoch": 1.0654863887302495, "grad_norm": 18.952045440673828, "learning_rate": 4.634740784185955e-06, "loss": 0.5423, "step": 1400 }, { "epoch": 1.0662478583666477, "grad_norm": 39.652442932128906, "learning_rate": 4.62866425555628e-06, "loss": 0.7428, "step": 1401 }, { "epoch": 1.067009328003046, "grad_norm": 26.903491973876953, "learning_rate": 4.622588278369598e-06, "loss": 0.622, "step": 1402 }, { "epoch": 1.0677707976394442, "grad_norm": 30.496828079223633, "learning_rate": 4.616512861648888e-06, "loss": 0.659, "step": 1403 }, { "epoch": 1.0685322672758424, "grad_norm": 25.840110778808594, "learning_rate": 4.610438014416299e-06, "loss": 0.6391, "step": 1404 }, { "epoch": 1.0692937369122406, "grad_norm": 25.98410987854004, "learning_rate": 4.604363745693129e-06, "loss": 0.6038, "step": 1405 }, { "epoch": 1.0700552065486388, "grad_norm": 35.958187103271484, "learning_rate": 4.598290064499825e-06, "loss": 0.8118, "step": 1406 }, { "epoch": 1.070816676185037, "grad_norm": 29.957862854003906, "learning_rate": 4.592216979855958e-06, "loss": 0.8409, "step": 1407 }, { "epoch": 1.0715781458214353, "grad_norm": 32.40114974975586, "learning_rate": 4.586144500780209e-06, "loss": 0.7027, "step": 1408 }, { "epoch": 1.0723396154578335, "grad_norm": 27.353778839111328, "learning_rate": 4.580072636290364e-06, "loss": 0.65, "step": 1409 }, { "epoch": 1.073101085094232, "grad_norm": 22.47332191467285, "learning_rate": 4.5740013954032985e-06, "loss": 0.5477, "step": 1410 }, { "epoch": 1.0738625547306302, "grad_norm": 26.15103530883789, "learning_rate": 4.5679307871349585e-06, "loss": 0.7346, "step": 1411 }, { "epoch": 1.0746240243670284, "grad_norm": 33.8211784362793, "learning_rate": 4.5618608205003475e-06, "loss": 0.6781, "step": 1412 }, { "epoch": 1.0753854940034266, "grad_norm": 34.45524597167969, "learning_rate": 4.555791504513524e-06, "loss": 0.8875, "step": 1413 }, { "epoch": 1.0761469636398249, "grad_norm": 25.472562789916992, "learning_rate": 4.5497228481875734e-06, "loss": 0.7432, "step": 1414 }, { "epoch": 1.076908433276223, "grad_norm": 25.834667205810547, "learning_rate": 4.543654860534607e-06, "loss": 0.7911, "step": 1415 }, { "epoch": 1.0776699029126213, "grad_norm": 32.063438415527344, "learning_rate": 4.537587550565735e-06, "loss": 0.8549, "step": 1416 }, { "epoch": 1.0784313725490196, "grad_norm": 27.36929702758789, "learning_rate": 4.531520927291073e-06, "loss": 0.6768, "step": 1417 }, { "epoch": 1.0791928421854178, "grad_norm": 26.77155113220215, "learning_rate": 4.525454999719706e-06, "loss": 0.5746, "step": 1418 }, { "epoch": 1.079954311821816, "grad_norm": 25.99591827392578, "learning_rate": 4.519389776859693e-06, "loss": 0.67, "step": 1419 }, { "epoch": 1.0807157814582145, "grad_norm": 22.180416107177734, "learning_rate": 4.513325267718039e-06, "loss": 0.7445, "step": 1420 }, { "epoch": 1.0814772510946127, "grad_norm": 36.45490646362305, "learning_rate": 4.507261481300699e-06, "loss": 0.903, "step": 1421 }, { "epoch": 1.082238720731011, "grad_norm": 21.906686782836914, "learning_rate": 4.501198426612547e-06, "loss": 0.5821, "step": 1422 }, { "epoch": 1.0830001903674091, "grad_norm": 34.67726516723633, "learning_rate": 4.495136112657371e-06, "loss": 0.9711, "step": 1423 }, { "epoch": 1.0837616600038074, "grad_norm": 29.7253360748291, "learning_rate": 4.489074548437866e-06, "loss": 0.8287, "step": 1424 }, { "epoch": 1.0845231296402056, "grad_norm": 25.894155502319336, "learning_rate": 4.483013742955605e-06, "loss": 0.6132, "step": 1425 }, { "epoch": 1.0852845992766038, "grad_norm": 33.52775573730469, "learning_rate": 4.47695370521104e-06, "loss": 0.7166, "step": 1426 }, { "epoch": 1.086046068913002, "grad_norm": 31.999540328979492, "learning_rate": 4.470894444203474e-06, "loss": 0.6902, "step": 1427 }, { "epoch": 1.0868075385494003, "grad_norm": 22.8083553314209, "learning_rate": 4.464835968931072e-06, "loss": 0.6289, "step": 1428 }, { "epoch": 1.0875690081857985, "grad_norm": 26.299516677856445, "learning_rate": 4.458778288390819e-06, "loss": 0.577, "step": 1429 }, { "epoch": 1.088330477822197, "grad_norm": 37.32050704956055, "learning_rate": 4.452721411578525e-06, "loss": 0.792, "step": 1430 }, { "epoch": 1.0890919474585952, "grad_norm": 27.388978958129883, "learning_rate": 4.446665347488803e-06, "loss": 0.5382, "step": 1431 }, { "epoch": 1.0898534170949934, "grad_norm": 31.408645629882812, "learning_rate": 4.440610105115065e-06, "loss": 0.7207, "step": 1432 }, { "epoch": 1.0906148867313916, "grad_norm": 23.884397506713867, "learning_rate": 4.434555693449498e-06, "loss": 0.5687, "step": 1433 }, { "epoch": 1.0913763563677898, "grad_norm": 22.35312843322754, "learning_rate": 4.428502121483056e-06, "loss": 0.5209, "step": 1434 }, { "epoch": 1.092137826004188, "grad_norm": 33.05360794067383, "learning_rate": 4.4224493982054504e-06, "loss": 0.8766, "step": 1435 }, { "epoch": 1.0928992956405863, "grad_norm": 31.350481033325195, "learning_rate": 4.416397532605124e-06, "loss": 0.8163, "step": 1436 }, { "epoch": 1.0936607652769845, "grad_norm": 27.195911407470703, "learning_rate": 4.4103465336692535e-06, "loss": 0.7951, "step": 1437 }, { "epoch": 1.0944222349133828, "grad_norm": 29.06740379333496, "learning_rate": 4.4042964103837235e-06, "loss": 0.6893, "step": 1438 }, { "epoch": 1.095183704549781, "grad_norm": 33.67143630981445, "learning_rate": 4.398247171733124e-06, "loss": 0.9014, "step": 1439 }, { "epoch": 1.0959451741861794, "grad_norm": 31.75952911376953, "learning_rate": 4.392198826700724e-06, "loss": 0.8854, "step": 1440 }, { "epoch": 1.0967066438225777, "grad_norm": 29.88365364074707, "learning_rate": 4.386151384268472e-06, "loss": 0.7514, "step": 1441 }, { "epoch": 1.0974681134589759, "grad_norm": 22.612136840820312, "learning_rate": 4.3801048534169676e-06, "loss": 0.6581, "step": 1442 }, { "epoch": 1.098229583095374, "grad_norm": 25.116926193237305, "learning_rate": 4.374059243125471e-06, "loss": 0.5853, "step": 1443 }, { "epoch": 1.0989910527317723, "grad_norm": 34.97810363769531, "learning_rate": 4.368014562371861e-06, "loss": 0.8078, "step": 1444 }, { "epoch": 1.0997525223681706, "grad_norm": 31.783151626586914, "learning_rate": 4.3619708201326415e-06, "loss": 0.83, "step": 1445 }, { "epoch": 1.1005139920045688, "grad_norm": 34.978248596191406, "learning_rate": 4.355928025382927e-06, "loss": 0.8357, "step": 1446 }, { "epoch": 1.101275461640967, "grad_norm": 33.985382080078125, "learning_rate": 4.349886187096417e-06, "loss": 0.7945, "step": 1447 }, { "epoch": 1.1020369312773652, "grad_norm": 27.279088973999023, "learning_rate": 4.343845314245396e-06, "loss": 0.6162, "step": 1448 }, { "epoch": 1.1027984009137635, "grad_norm": 28.585399627685547, "learning_rate": 4.337805415800712e-06, "loss": 0.7662, "step": 1449 }, { "epoch": 1.103559870550162, "grad_norm": 36.59597396850586, "learning_rate": 4.331766500731771e-06, "loss": 0.5001, "step": 1450 }, { "epoch": 1.1043213401865601, "grad_norm": 29.027137756347656, "learning_rate": 4.32572857800651e-06, "loss": 0.661, "step": 1451 }, { "epoch": 1.1050828098229584, "grad_norm": 30.55579948425293, "learning_rate": 4.319691656591401e-06, "loss": 0.7415, "step": 1452 }, { "epoch": 1.1058442794593566, "grad_norm": 26.680347442626953, "learning_rate": 4.313655745451422e-06, "loss": 0.6562, "step": 1453 }, { "epoch": 1.1066057490957548, "grad_norm": 28.822954177856445, "learning_rate": 4.3076208535500575e-06, "loss": 0.6864, "step": 1454 }, { "epoch": 1.107367218732153, "grad_norm": 28.9395751953125, "learning_rate": 4.301586989849272e-06, "loss": 0.6809, "step": 1455 }, { "epoch": 1.1081286883685513, "grad_norm": 25.72883415222168, "learning_rate": 4.295554163309509e-06, "loss": 0.7297, "step": 1456 }, { "epoch": 1.1088901580049495, "grad_norm": 32.15998840332031, "learning_rate": 4.289522382889664e-06, "loss": 0.7653, "step": 1457 }, { "epoch": 1.1096516276413477, "grad_norm": 23.909706115722656, "learning_rate": 4.28349165754709e-06, "loss": 0.7175, "step": 1458 }, { "epoch": 1.110413097277746, "grad_norm": 27.97369384765625, "learning_rate": 4.277461996237563e-06, "loss": 0.8097, "step": 1459 }, { "epoch": 1.1111745669141444, "grad_norm": 52.83234405517578, "learning_rate": 4.271433407915283e-06, "loss": 1.0129, "step": 1460 }, { "epoch": 1.1119360365505426, "grad_norm": 27.569765090942383, "learning_rate": 4.265405901532857e-06, "loss": 0.5832, "step": 1461 }, { "epoch": 1.1126975061869409, "grad_norm": 29.594486236572266, "learning_rate": 4.259379486041287e-06, "loss": 0.5924, "step": 1462 }, { "epoch": 1.113458975823339, "grad_norm": 31.912498474121094, "learning_rate": 4.25335417038995e-06, "loss": 0.8122, "step": 1463 }, { "epoch": 1.1142204454597373, "grad_norm": 23.730451583862305, "learning_rate": 4.2473299635265906e-06, "loss": 0.6029, "step": 1464 }, { "epoch": 1.1149819150961355, "grad_norm": 33.216796875, "learning_rate": 4.241306874397314e-06, "loss": 0.8299, "step": 1465 }, { "epoch": 1.1157433847325338, "grad_norm": 34.02432632446289, "learning_rate": 4.235284911946557e-06, "loss": 0.8686, "step": 1466 }, { "epoch": 1.116504854368932, "grad_norm": 34.81566619873047, "learning_rate": 4.229264085117086e-06, "loss": 0.8477, "step": 1467 }, { "epoch": 1.1172663240053302, "grad_norm": 34.48906326293945, "learning_rate": 4.223244402849981e-06, "loss": 0.794, "step": 1468 }, { "epoch": 1.1180277936417284, "grad_norm": 25.180116653442383, "learning_rate": 4.217225874084628e-06, "loss": 0.7513, "step": 1469 }, { "epoch": 1.1187892632781269, "grad_norm": 27.25897789001465, "learning_rate": 4.2112085077586875e-06, "loss": 0.6357, "step": 1470 }, { "epoch": 1.1195507329145251, "grad_norm": 25.58527946472168, "learning_rate": 4.205192312808104e-06, "loss": 0.5776, "step": 1471 }, { "epoch": 1.1203122025509233, "grad_norm": 27.980712890625, "learning_rate": 4.199177298167081e-06, "loss": 0.6772, "step": 1472 }, { "epoch": 1.1210736721873216, "grad_norm": 24.157320022583008, "learning_rate": 4.193163472768067e-06, "loss": 0.6443, "step": 1473 }, { "epoch": 1.1218351418237198, "grad_norm": 26.085046768188477, "learning_rate": 4.1871508455417434e-06, "loss": 0.8099, "step": 1474 }, { "epoch": 1.122596611460118, "grad_norm": 30.226167678833008, "learning_rate": 4.181139425417015e-06, "loss": 0.6938, "step": 1475 }, { "epoch": 1.1233580810965162, "grad_norm": 30.745046615600586, "learning_rate": 4.175129221320993e-06, "loss": 0.7864, "step": 1476 }, { "epoch": 1.1241195507329145, "grad_norm": 29.392539978027344, "learning_rate": 4.169120242178986e-06, "loss": 0.8462, "step": 1477 }, { "epoch": 1.1248810203693127, "grad_norm": 25.927181243896484, "learning_rate": 4.163112496914475e-06, "loss": 0.6139, "step": 1478 }, { "epoch": 1.125642490005711, "grad_norm": 27.550506591796875, "learning_rate": 4.157105994449115e-06, "loss": 0.6812, "step": 1479 }, { "epoch": 1.1264039596421092, "grad_norm": 30.582822799682617, "learning_rate": 4.151100743702719e-06, "loss": 0.7502, "step": 1480 }, { "epoch": 1.1271654292785076, "grad_norm": 35.46127700805664, "learning_rate": 4.145096753593232e-06, "loss": 0.78, "step": 1481 }, { "epoch": 1.1279268989149058, "grad_norm": 20.126033782958984, "learning_rate": 4.1390940330367305e-06, "loss": 0.6424, "step": 1482 }, { "epoch": 1.128688368551304, "grad_norm": 23.577369689941406, "learning_rate": 4.1330925909474105e-06, "loss": 0.5585, "step": 1483 }, { "epoch": 1.1294498381877023, "grad_norm": 34.41134262084961, "learning_rate": 4.1270924362375656e-06, "loss": 0.8326, "step": 1484 }, { "epoch": 1.1302113078241005, "grad_norm": 30.686939239501953, "learning_rate": 4.121093577817575e-06, "loss": 0.7246, "step": 1485 }, { "epoch": 1.1309727774604987, "grad_norm": 31.418886184692383, "learning_rate": 4.115096024595895e-06, "loss": 0.7177, "step": 1486 }, { "epoch": 1.131734247096897, "grad_norm": 25.855369567871094, "learning_rate": 4.109099785479047e-06, "loss": 0.6648, "step": 1487 }, { "epoch": 1.1324957167332952, "grad_norm": 25.41705322265625, "learning_rate": 4.103104869371597e-06, "loss": 0.7566, "step": 1488 }, { "epoch": 1.1332571863696934, "grad_norm": 27.526826858520508, "learning_rate": 4.0971112851761476e-06, "loss": 0.7355, "step": 1489 }, { "epoch": 1.1340186560060919, "grad_norm": 27.98457908630371, "learning_rate": 4.0911190417933205e-06, "loss": 0.692, "step": 1490 }, { "epoch": 1.13478012564249, "grad_norm": 28.406068801879883, "learning_rate": 4.085128148121753e-06, "loss": 0.5555, "step": 1491 }, { "epoch": 1.1355415952788883, "grad_norm": 41.409358978271484, "learning_rate": 4.079138613058072e-06, "loss": 0.7789, "step": 1492 }, { "epoch": 1.1363030649152865, "grad_norm": 38.29566955566406, "learning_rate": 4.073150445496888e-06, "loss": 0.7782, "step": 1493 }, { "epoch": 1.1370645345516848, "grad_norm": 31.088985443115234, "learning_rate": 4.067163654330784e-06, "loss": 0.7208, "step": 1494 }, { "epoch": 1.137826004188083, "grad_norm": 28.536766052246094, "learning_rate": 4.061178248450295e-06, "loss": 0.5844, "step": 1495 }, { "epoch": 1.1385874738244812, "grad_norm": 37.34573745727539, "learning_rate": 4.055194236743901e-06, "loss": 1.1118, "step": 1496 }, { "epoch": 1.1393489434608794, "grad_norm": 27.20584487915039, "learning_rate": 4.049211628098009e-06, "loss": 0.6863, "step": 1497 }, { "epoch": 1.1401104130972777, "grad_norm": 21.4528751373291, "learning_rate": 4.043230431396948e-06, "loss": 0.4839, "step": 1498 }, { "epoch": 1.140871882733676, "grad_norm": 33.37261199951172, "learning_rate": 4.037250655522946e-06, "loss": 0.8318, "step": 1499 }, { "epoch": 1.1416333523700741, "grad_norm": 35.733123779296875, "learning_rate": 4.031272309356119e-06, "loss": 0.8357, "step": 1500 }, { "epoch": 1.1423948220064726, "grad_norm": 28.71573829650879, "learning_rate": 4.025295401774463e-06, "loss": 0.6477, "step": 1501 }, { "epoch": 1.1431562916428708, "grad_norm": 25.098535537719727, "learning_rate": 4.01931994165384e-06, "loss": 0.5957, "step": 1502 }, { "epoch": 1.143917761279269, "grad_norm": 33.505836486816406, "learning_rate": 4.013345937867959e-06, "loss": 0.6947, "step": 1503 }, { "epoch": 1.1446792309156673, "grad_norm": 30.04947853088379, "learning_rate": 4.007373399288364e-06, "loss": 0.7854, "step": 1504 }, { "epoch": 1.1454407005520655, "grad_norm": 26.88728904724121, "learning_rate": 4.00140233478443e-06, "loss": 0.4959, "step": 1505 }, { "epoch": 1.1462021701884637, "grad_norm": 38.54801940917969, "learning_rate": 3.995432753223338e-06, "loss": 0.8951, "step": 1506 }, { "epoch": 1.146963639824862, "grad_norm": 39.258689880371094, "learning_rate": 3.989464663470068e-06, "loss": 0.7519, "step": 1507 }, { "epoch": 1.1477251094612602, "grad_norm": 29.316484451293945, "learning_rate": 3.983498074387381e-06, "loss": 0.5598, "step": 1508 }, { "epoch": 1.1484865790976584, "grad_norm": 27.5557804107666, "learning_rate": 3.977532994835818e-06, "loss": 0.5955, "step": 1509 }, { "epoch": 1.1492480487340568, "grad_norm": 22.647356033325195, "learning_rate": 3.9715694336736705e-06, "loss": 0.4456, "step": 1510 }, { "epoch": 1.150009518370455, "grad_norm": 30.45392417907715, "learning_rate": 3.96560739975698e-06, "loss": 0.7109, "step": 1511 }, { "epoch": 1.1507709880068533, "grad_norm": 31.624401092529297, "learning_rate": 3.959646901939515e-06, "loss": 0.7487, "step": 1512 }, { "epoch": 1.1515324576432515, "grad_norm": 33.577571868896484, "learning_rate": 3.953687949072769e-06, "loss": 0.8535, "step": 1513 }, { "epoch": 1.1522939272796497, "grad_norm": 24.781082153320312, "learning_rate": 3.9477305500059356e-06, "loss": 0.5593, "step": 1514 }, { "epoch": 1.153055396916048, "grad_norm": 35.7469482421875, "learning_rate": 3.941774713585903e-06, "loss": 0.8245, "step": 1515 }, { "epoch": 1.1538168665524462, "grad_norm": 31.718555450439453, "learning_rate": 3.9358204486572435e-06, "loss": 0.6243, "step": 1516 }, { "epoch": 1.1545783361888444, "grad_norm": 35.048465728759766, "learning_rate": 3.929867764062187e-06, "loss": 1.0063, "step": 1517 }, { "epoch": 1.1553398058252426, "grad_norm": 28.552074432373047, "learning_rate": 3.923916668640625e-06, "loss": 0.7059, "step": 1518 }, { "epoch": 1.1561012754616409, "grad_norm": 38.085662841796875, "learning_rate": 3.917967171230078e-06, "loss": 0.935, "step": 1519 }, { "epoch": 1.156862745098039, "grad_norm": 27.50920295715332, "learning_rate": 3.912019280665708e-06, "loss": 0.7094, "step": 1520 }, { "epoch": 1.1576242147344376, "grad_norm": 35.8309440612793, "learning_rate": 3.90607300578028e-06, "loss": 0.6515, "step": 1521 }, { "epoch": 1.1583856843708358, "grad_norm": 23.801332473754883, "learning_rate": 3.900128355404163e-06, "loss": 0.5345, "step": 1522 }, { "epoch": 1.159147154007234, "grad_norm": 30.352127075195312, "learning_rate": 3.8941853383653125e-06, "loss": 0.5819, "step": 1523 }, { "epoch": 1.1599086236436322, "grad_norm": 28.999980926513672, "learning_rate": 3.888243963489262e-06, "loss": 0.7428, "step": 1524 }, { "epoch": 1.1606700932800305, "grad_norm": 29.182889938354492, "learning_rate": 3.882304239599102e-06, "loss": 0.5231, "step": 1525 }, { "epoch": 1.1614315629164287, "grad_norm": 29.404163360595703, "learning_rate": 3.876366175515472e-06, "loss": 0.731, "step": 1526 }, { "epoch": 1.162193032552827, "grad_norm": 31.57741928100586, "learning_rate": 3.870429780056551e-06, "loss": 0.781, "step": 1527 }, { "epoch": 1.1629545021892251, "grad_norm": 24.528194427490234, "learning_rate": 3.864495062038033e-06, "loss": 0.6924, "step": 1528 }, { "epoch": 1.1637159718256234, "grad_norm": 31.90503692626953, "learning_rate": 3.858562030273127e-06, "loss": 0.8175, "step": 1529 }, { "epoch": 1.1644774414620218, "grad_norm": 30.26751136779785, "learning_rate": 3.852630693572534e-06, "loss": 0.7417, "step": 1530 }, { "epoch": 1.16523891109842, "grad_norm": 31.814529418945312, "learning_rate": 3.846701060744442e-06, "loss": 0.7372, "step": 1531 }, { "epoch": 1.1660003807348183, "grad_norm": 31.711366653442383, "learning_rate": 3.840773140594503e-06, "loss": 0.8018, "step": 1532 }, { "epoch": 1.1667618503712165, "grad_norm": 30.717601776123047, "learning_rate": 3.834846941925832e-06, "loss": 0.6855, "step": 1533 }, { "epoch": 1.1675233200076147, "grad_norm": 31.86241912841797, "learning_rate": 3.828922473538978e-06, "loss": 0.7614, "step": 1534 }, { "epoch": 1.168284789644013, "grad_norm": 23.519512176513672, "learning_rate": 3.822999744231933e-06, "loss": 0.539, "step": 1535 }, { "epoch": 1.1690462592804112, "grad_norm": 29.818391799926758, "learning_rate": 3.817078762800099e-06, "loss": 0.7589, "step": 1536 }, { "epoch": 1.1698077289168094, "grad_norm": 29.188121795654297, "learning_rate": 3.811159538036282e-06, "loss": 0.5358, "step": 1537 }, { "epoch": 1.1705691985532076, "grad_norm": 29.57836151123047, "learning_rate": 3.8052420787306785e-06, "loss": 0.6442, "step": 1538 }, { "epoch": 1.1713306681896059, "grad_norm": 31.207075119018555, "learning_rate": 3.799326393670872e-06, "loss": 0.7802, "step": 1539 }, { "epoch": 1.172092137826004, "grad_norm": 22.89545249938965, "learning_rate": 3.793412491641799e-06, "loss": 0.7152, "step": 1540 }, { "epoch": 1.1728536074624025, "grad_norm": 28.425622940063477, "learning_rate": 3.7875003814257553e-06, "loss": 0.6614, "step": 1541 }, { "epoch": 1.1736150770988008, "grad_norm": 27.53705596923828, "learning_rate": 3.781590071802378e-06, "loss": 0.6771, "step": 1542 }, { "epoch": 1.174376546735199, "grad_norm": 25.33406639099121, "learning_rate": 3.775681571548622e-06, "loss": 0.678, "step": 1543 }, { "epoch": 1.1751380163715972, "grad_norm": 24.779775619506836, "learning_rate": 3.7697748894387607e-06, "loss": 0.637, "step": 1544 }, { "epoch": 1.1758994860079954, "grad_norm": 27.581829071044922, "learning_rate": 3.7638700342443673e-06, "loss": 0.8448, "step": 1545 }, { "epoch": 1.1766609556443937, "grad_norm": 29.22455406188965, "learning_rate": 3.7579670147343037e-06, "loss": 0.6459, "step": 1546 }, { "epoch": 1.1774224252807919, "grad_norm": 28.1949405670166, "learning_rate": 3.752065839674701e-06, "loss": 0.7035, "step": 1547 }, { "epoch": 1.17818389491719, "grad_norm": 24.18442153930664, "learning_rate": 3.746166517828954e-06, "loss": 0.7038, "step": 1548 }, { "epoch": 1.1789453645535883, "grad_norm": 38.9047737121582, "learning_rate": 3.740269057957705e-06, "loss": 0.8302, "step": 1549 }, { "epoch": 1.1797068341899868, "grad_norm": 29.040796279907227, "learning_rate": 3.7343734688188328e-06, "loss": 0.7768, "step": 1550 }, { "epoch": 1.180468303826385, "grad_norm": 23.671512603759766, "learning_rate": 3.7284797591674355e-06, "loss": 0.7409, "step": 1551 }, { "epoch": 1.1812297734627832, "grad_norm": 25.06656837463379, "learning_rate": 3.72258793775582e-06, "loss": 0.5625, "step": 1552 }, { "epoch": 1.1819912430991815, "grad_norm": 34.21515655517578, "learning_rate": 3.7166980133334928e-06, "loss": 0.9147, "step": 1553 }, { "epoch": 1.1827527127355797, "grad_norm": 21.93282699584961, "learning_rate": 3.7108099946471422e-06, "loss": 0.62, "step": 1554 }, { "epoch": 1.183514182371978, "grad_norm": 29.166244506835938, "learning_rate": 3.7049238904406205e-06, "loss": 0.6489, "step": 1555 }, { "epoch": 1.1842756520083761, "grad_norm": 34.56402587890625, "learning_rate": 3.699039709454944e-06, "loss": 0.814, "step": 1556 }, { "epoch": 1.1850371216447744, "grad_norm": 25.795196533203125, "learning_rate": 3.693157460428273e-06, "loss": 0.6034, "step": 1557 }, { "epoch": 1.1857985912811726, "grad_norm": 44.57625961303711, "learning_rate": 3.6872771520958926e-06, "loss": 0.8259, "step": 1558 }, { "epoch": 1.1865600609175708, "grad_norm": 32.11421203613281, "learning_rate": 3.6813987931902116e-06, "loss": 0.7835, "step": 1559 }, { "epoch": 1.187321530553969, "grad_norm": 27.94780731201172, "learning_rate": 3.67552239244074e-06, "loss": 0.7619, "step": 1560 }, { "epoch": 1.1880830001903675, "grad_norm": 32.07542419433594, "learning_rate": 3.6696479585740864e-06, "loss": 0.8366, "step": 1561 }, { "epoch": 1.1888444698267657, "grad_norm": 27.724348068237305, "learning_rate": 3.6637755003139295e-06, "loss": 0.6817, "step": 1562 }, { "epoch": 1.189605939463164, "grad_norm": 28.572914123535156, "learning_rate": 3.657905026381019e-06, "loss": 0.6858, "step": 1563 }, { "epoch": 1.1903674090995622, "grad_norm": 55.64738845825195, "learning_rate": 3.65203654549316e-06, "loss": 0.8185, "step": 1564 }, { "epoch": 1.1911288787359604, "grad_norm": 34.84992980957031, "learning_rate": 3.6461700663651944e-06, "loss": 0.7621, "step": 1565 }, { "epoch": 1.1918903483723586, "grad_norm": 26.228931427001953, "learning_rate": 3.640305597708992e-06, "loss": 0.6616, "step": 1566 }, { "epoch": 1.1926518180087569, "grad_norm": 24.440567016601562, "learning_rate": 3.6344431482334356e-06, "loss": 0.799, "step": 1567 }, { "epoch": 1.193413287645155, "grad_norm": 22.929447174072266, "learning_rate": 3.628582726644414e-06, "loss": 0.5972, "step": 1568 }, { "epoch": 1.1941747572815533, "grad_norm": 29.916950225830078, "learning_rate": 3.6227243416448023e-06, "loss": 0.7779, "step": 1569 }, { "epoch": 1.1949362269179518, "grad_norm": 29.62129783630371, "learning_rate": 3.6168680019344484e-06, "loss": 0.7305, "step": 1570 }, { "epoch": 1.19569769655435, "grad_norm": 21.802913665771484, "learning_rate": 3.6110137162101646e-06, "loss": 0.5038, "step": 1571 }, { "epoch": 1.1964591661907482, "grad_norm": 25.729015350341797, "learning_rate": 3.6051614931657154e-06, "loss": 0.6479, "step": 1572 }, { "epoch": 1.1972206358271464, "grad_norm": 26.470075607299805, "learning_rate": 3.5993113414918017e-06, "loss": 0.7057, "step": 1573 }, { "epoch": 1.1979821054635447, "grad_norm": 26.32611083984375, "learning_rate": 3.593463269876043e-06, "loss": 0.5745, "step": 1574 }, { "epoch": 1.198743575099943, "grad_norm": 36.525150299072266, "learning_rate": 3.5876172870029768e-06, "loss": 0.7234, "step": 1575 }, { "epoch": 1.1995050447363411, "grad_norm": 35.58689498901367, "learning_rate": 3.5817734015540368e-06, "loss": 0.8185, "step": 1576 }, { "epoch": 1.2002665143727393, "grad_norm": 29.15963363647461, "learning_rate": 3.575931622207538e-06, "loss": 0.7797, "step": 1577 }, { "epoch": 1.2010279840091376, "grad_norm": 32.886985778808594, "learning_rate": 3.5700919576386707e-06, "loss": 0.7744, "step": 1578 }, { "epoch": 1.2017894536455358, "grad_norm": 26.013147354125977, "learning_rate": 3.564254416519488e-06, "loss": 0.7642, "step": 1579 }, { "epoch": 1.202550923281934, "grad_norm": 31.868709564208984, "learning_rate": 3.5584190075188853e-06, "loss": 0.9576, "step": 1580 }, { "epoch": 1.2033123929183325, "grad_norm": 26.204513549804688, "learning_rate": 3.552585739302591e-06, "loss": 0.7331, "step": 1581 }, { "epoch": 1.2040738625547307, "grad_norm": 25.806148529052734, "learning_rate": 3.546754620533156e-06, "loss": 0.6407, "step": 1582 }, { "epoch": 1.204835332191129, "grad_norm": 25.98582649230957, "learning_rate": 3.540925659869942e-06, "loss": 0.6004, "step": 1583 }, { "epoch": 1.2055968018275272, "grad_norm": 26.506555557250977, "learning_rate": 3.5350988659691034e-06, "loss": 0.7565, "step": 1584 }, { "epoch": 1.2063582714639254, "grad_norm": 30.192405700683594, "learning_rate": 3.5292742474835738e-06, "loss": 0.654, "step": 1585 }, { "epoch": 1.2071197411003236, "grad_norm": 30.088260650634766, "learning_rate": 3.523451813063062e-06, "loss": 0.7091, "step": 1586 }, { "epoch": 1.2078812107367218, "grad_norm": 27.887300491333008, "learning_rate": 3.517631571354029e-06, "loss": 0.6445, "step": 1587 }, { "epoch": 1.20864268037312, "grad_norm": 28.69025230407715, "learning_rate": 3.5118135309996837e-06, "loss": 0.7182, "step": 1588 }, { "epoch": 1.2094041500095183, "grad_norm": 32.164451599121094, "learning_rate": 3.5059977006399598e-06, "loss": 0.7542, "step": 1589 }, { "epoch": 1.2101656196459167, "grad_norm": 35.95284652709961, "learning_rate": 3.5001840889115157e-06, "loss": 0.938, "step": 1590 }, { "epoch": 1.210927089282315, "grad_norm": 24.52789306640625, "learning_rate": 3.494372704447712e-06, "loss": 0.5769, "step": 1591 }, { "epoch": 1.2116885589187132, "grad_norm": 36.9116325378418, "learning_rate": 3.4885635558786e-06, "loss": 0.7539, "step": 1592 }, { "epoch": 1.2124500285551114, "grad_norm": 27.540515899658203, "learning_rate": 3.482756651830913e-06, "loss": 0.7939, "step": 1593 }, { "epoch": 1.2132114981915096, "grad_norm": 28.228578567504883, "learning_rate": 3.4769520009280524e-06, "loss": 0.7848, "step": 1594 }, { "epoch": 1.2139729678279079, "grad_norm": 25.078903198242188, "learning_rate": 3.471149611790072e-06, "loss": 0.7105, "step": 1595 }, { "epoch": 1.214734437464306, "grad_norm": 30.21971893310547, "learning_rate": 3.465349493033662e-06, "loss": 0.5797, "step": 1596 }, { "epoch": 1.2154959071007043, "grad_norm": 31.039281845092773, "learning_rate": 3.4595516532721523e-06, "loss": 0.8083, "step": 1597 }, { "epoch": 1.2162573767371025, "grad_norm": 34.632877349853516, "learning_rate": 3.4537561011154787e-06, "loss": 0.8025, "step": 1598 }, { "epoch": 1.2170188463735008, "grad_norm": 27.43385124206543, "learning_rate": 3.4479628451701834e-06, "loss": 0.684, "step": 1599 }, { "epoch": 1.217780316009899, "grad_norm": 28.920392990112305, "learning_rate": 3.4421718940393965e-06, "loss": 0.7268, "step": 1600 }, { "epoch": 1.2185417856462974, "grad_norm": 22.443618774414062, "learning_rate": 3.4363832563228306e-06, "loss": 0.625, "step": 1601 }, { "epoch": 1.2193032552826957, "grad_norm": 28.5852108001709, "learning_rate": 3.4305969406167566e-06, "loss": 0.8282, "step": 1602 }, { "epoch": 1.220064724919094, "grad_norm": 22.95713233947754, "learning_rate": 3.4248129555140024e-06, "loss": 0.5918, "step": 1603 }, { "epoch": 1.2208261945554921, "grad_norm": 25.616674423217773, "learning_rate": 3.4190313096039296e-06, "loss": 0.5247, "step": 1604 }, { "epoch": 1.2215876641918904, "grad_norm": 34.38652420043945, "learning_rate": 3.413252011472431e-06, "loss": 0.858, "step": 1605 }, { "epoch": 1.2223491338282886, "grad_norm": 33.26419448852539, "learning_rate": 3.40747506970191e-06, "loss": 0.7209, "step": 1606 }, { "epoch": 1.2231106034646868, "grad_norm": 27.42671775817871, "learning_rate": 3.4017004928712726e-06, "loss": 0.8872, "step": 1607 }, { "epoch": 1.223872073101085, "grad_norm": 33.50828170776367, "learning_rate": 3.39592828955591e-06, "loss": 0.8912, "step": 1608 }, { "epoch": 1.2246335427374833, "grad_norm": 24.625503540039062, "learning_rate": 3.390158468327691e-06, "loss": 0.5305, "step": 1609 }, { "epoch": 1.2253950123738817, "grad_norm": 31.81485939025879, "learning_rate": 3.3843910377549488e-06, "loss": 0.6929, "step": 1610 }, { "epoch": 1.22615648201028, "grad_norm": 27.919307708740234, "learning_rate": 3.3786260064024582e-06, "loss": 0.6368, "step": 1611 }, { "epoch": 1.2269179516466782, "grad_norm": 26.418922424316406, "learning_rate": 3.372863382831444e-06, "loss": 0.7177, "step": 1612 }, { "epoch": 1.2276794212830764, "grad_norm": 26.7669620513916, "learning_rate": 3.367103175599543e-06, "loss": 0.7483, "step": 1613 }, { "epoch": 1.2284408909194746, "grad_norm": 22.196243286132812, "learning_rate": 3.3613453932608126e-06, "loss": 0.5872, "step": 1614 }, { "epoch": 1.2292023605558728, "grad_norm": 26.478099822998047, "learning_rate": 3.3555900443657e-06, "loss": 0.698, "step": 1615 }, { "epoch": 1.229963830192271, "grad_norm": 31.410436630249023, "learning_rate": 3.349837137461051e-06, "loss": 1.0238, "step": 1616 }, { "epoch": 1.2307252998286693, "grad_norm": 31.702730178833008, "learning_rate": 3.344086681090074e-06, "loss": 0.7804, "step": 1617 }, { "epoch": 1.2314867694650675, "grad_norm": 29.988819122314453, "learning_rate": 3.338338683792345e-06, "loss": 0.8951, "step": 1618 }, { "epoch": 1.2322482391014657, "grad_norm": 29.898778915405273, "learning_rate": 3.332593154103781e-06, "loss": 0.7516, "step": 1619 }, { "epoch": 1.233009708737864, "grad_norm": 27.284597396850586, "learning_rate": 3.3268501005566435e-06, "loss": 0.7139, "step": 1620 }, { "epoch": 1.2337711783742624, "grad_norm": 28.379440307617188, "learning_rate": 3.3211095316795117e-06, "loss": 0.7116, "step": 1621 }, { "epoch": 1.2345326480106606, "grad_norm": 22.33309555053711, "learning_rate": 3.3153714559972737e-06, "loss": 0.421, "step": 1622 }, { "epoch": 1.2352941176470589, "grad_norm": 24.770538330078125, "learning_rate": 3.309635882031122e-06, "loss": 0.5567, "step": 1623 }, { "epoch": 1.236055587283457, "grad_norm": 27.292123794555664, "learning_rate": 3.303902818298525e-06, "loss": 0.6962, "step": 1624 }, { "epoch": 1.2368170569198553, "grad_norm": 24.22323989868164, "learning_rate": 3.2981722733132304e-06, "loss": 0.5146, "step": 1625 }, { "epoch": 1.2375785265562536, "grad_norm": 26.400760650634766, "learning_rate": 3.2924442555852377e-06, "loss": 0.75, "step": 1626 }, { "epoch": 1.2383399961926518, "grad_norm": 26.19630241394043, "learning_rate": 3.2867187736208057e-06, "loss": 0.673, "step": 1627 }, { "epoch": 1.23910146582905, "grad_norm": 22.45516014099121, "learning_rate": 3.2809958359224153e-06, "loss": 0.4988, "step": 1628 }, { "epoch": 1.2398629354654482, "grad_norm": 27.836393356323242, "learning_rate": 3.275275450988776e-06, "loss": 0.6584, "step": 1629 }, { "epoch": 1.2406244051018467, "grad_norm": 26.4266357421875, "learning_rate": 3.269557627314802e-06, "loss": 0.7084, "step": 1630 }, { "epoch": 1.2413858747382447, "grad_norm": 32.049537658691406, "learning_rate": 3.2638423733916103e-06, "loss": 0.852, "step": 1631 }, { "epoch": 1.2421473443746431, "grad_norm": 30.954370498657227, "learning_rate": 3.2581296977064948e-06, "loss": 0.6864, "step": 1632 }, { "epoch": 1.2429088140110414, "grad_norm": 38.59880828857422, "learning_rate": 3.2524196087429232e-06, "loss": 0.9067, "step": 1633 }, { "epoch": 1.2436702836474396, "grad_norm": 30.154590606689453, "learning_rate": 3.246712114980526e-06, "loss": 0.7411, "step": 1634 }, { "epoch": 1.2444317532838378, "grad_norm": 26.712778091430664, "learning_rate": 3.2410072248950723e-06, "loss": 0.6456, "step": 1635 }, { "epoch": 1.245193222920236, "grad_norm": 38.5605583190918, "learning_rate": 3.235304946958471e-06, "loss": 0.8441, "step": 1636 }, { "epoch": 1.2459546925566343, "grad_norm": 29.960294723510742, "learning_rate": 3.229605289638746e-06, "loss": 0.5755, "step": 1637 }, { "epoch": 1.2467161621930325, "grad_norm": 27.653017044067383, "learning_rate": 3.223908261400037e-06, "loss": 0.6706, "step": 1638 }, { "epoch": 1.2474776318294307, "grad_norm": 26.435380935668945, "learning_rate": 3.2182138707025733e-06, "loss": 0.7156, "step": 1639 }, { "epoch": 1.248239101465829, "grad_norm": 34.11128234863281, "learning_rate": 3.2125221260026696e-06, "loss": 0.8128, "step": 1640 }, { "epoch": 1.2490005711022274, "grad_norm": 25.915218353271484, "learning_rate": 3.20683303575271e-06, "loss": 0.6808, "step": 1641 }, { "epoch": 1.2497620407386256, "grad_norm": 22.69007110595703, "learning_rate": 3.20114660840114e-06, "loss": 0.5157, "step": 1642 }, { "epoch": 1.2505235103750239, "grad_norm": 28.969921112060547, "learning_rate": 3.195462852392447e-06, "loss": 0.6484, "step": 1643 }, { "epoch": 1.251284980011422, "grad_norm": 27.462810516357422, "learning_rate": 3.189781776167151e-06, "loss": 0.8242, "step": 1644 }, { "epoch": 1.2520464496478203, "grad_norm": 23.542844772338867, "learning_rate": 3.1841033881617966e-06, "loss": 0.6131, "step": 1645 }, { "epoch": 1.2528079192842185, "grad_norm": 25.50222396850586, "learning_rate": 3.178427696808935e-06, "loss": 0.6792, "step": 1646 }, { "epoch": 1.2535693889206168, "grad_norm": 27.3731689453125, "learning_rate": 3.172754710537109e-06, "loss": 0.658, "step": 1647 }, { "epoch": 1.254330858557015, "grad_norm": 30.30860137939453, "learning_rate": 3.1670844377708465e-06, "loss": 0.6945, "step": 1648 }, { "epoch": 1.2550923281934132, "grad_norm": 28.736616134643555, "learning_rate": 3.161416886930648e-06, "loss": 0.7103, "step": 1649 }, { "epoch": 1.2558537978298117, "grad_norm": 25.66908073425293, "learning_rate": 3.1557520664329722e-06, "loss": 0.7284, "step": 1650 }, { "epoch": 1.2566152674662097, "grad_norm": 28.471736907958984, "learning_rate": 3.1500899846902173e-06, "loss": 0.7219, "step": 1651 }, { "epoch": 1.257376737102608, "grad_norm": 39.2846565246582, "learning_rate": 3.144430650110718e-06, "loss": 0.9714, "step": 1652 }, { "epoch": 1.2581382067390063, "grad_norm": 32.47160720825195, "learning_rate": 3.138774071098734e-06, "loss": 0.8451, "step": 1653 }, { "epoch": 1.2588996763754046, "grad_norm": 26.18456268310547, "learning_rate": 3.1331202560544247e-06, "loss": 0.7234, "step": 1654 }, { "epoch": 1.2596611460118028, "grad_norm": 24.949220657348633, "learning_rate": 3.1274692133738483e-06, "loss": 0.5221, "step": 1655 }, { "epoch": 1.260422615648201, "grad_norm": 34.268619537353516, "learning_rate": 3.1218209514489493e-06, "loss": 0.8612, "step": 1656 }, { "epoch": 1.2611840852845992, "grad_norm": 26.842496871948242, "learning_rate": 3.116175478667539e-06, "loss": 0.6996, "step": 1657 }, { "epoch": 1.2619455549209975, "grad_norm": 28.278871536254883, "learning_rate": 3.110532803413286e-06, "loss": 0.7824, "step": 1658 }, { "epoch": 1.2627070245573957, "grad_norm": 30.750934600830078, "learning_rate": 3.1048929340657057e-06, "loss": 0.6337, "step": 1659 }, { "epoch": 1.263468494193794, "grad_norm": 27.03371810913086, "learning_rate": 3.0992558790001493e-06, "loss": 0.6441, "step": 1660 }, { "epoch": 1.2642299638301924, "grad_norm": 27.08733367919922, "learning_rate": 3.0936216465877865e-06, "loss": 0.6429, "step": 1661 }, { "epoch": 1.2649914334665906, "grad_norm": 26.30339241027832, "learning_rate": 3.0879902451955935e-06, "loss": 0.7312, "step": 1662 }, { "epoch": 1.2657529031029888, "grad_norm": 22.198081970214844, "learning_rate": 3.082361683186344e-06, "loss": 0.7181, "step": 1663 }, { "epoch": 1.266514372739387, "grad_norm": 29.07954216003418, "learning_rate": 3.076735968918596e-06, "loss": 0.6804, "step": 1664 }, { "epoch": 1.2672758423757853, "grad_norm": 24.36330795288086, "learning_rate": 3.0711131107466795e-06, "loss": 0.6268, "step": 1665 }, { "epoch": 1.2680373120121835, "grad_norm": 28.795000076293945, "learning_rate": 3.0654931170206782e-06, "loss": 0.7904, "step": 1666 }, { "epoch": 1.2687987816485817, "grad_norm": 31.75333595275879, "learning_rate": 3.0598759960864265e-06, "loss": 0.6871, "step": 1667 }, { "epoch": 1.26956025128498, "grad_norm": 33.90879821777344, "learning_rate": 3.0542617562854928e-06, "loss": 0.7308, "step": 1668 }, { "epoch": 1.2703217209213782, "grad_norm": 31.590106964111328, "learning_rate": 3.0486504059551636e-06, "loss": 0.7083, "step": 1669 }, { "epoch": 1.2710831905577766, "grad_norm": 23.969661712646484, "learning_rate": 3.0430419534284356e-06, "loss": 0.4843, "step": 1670 }, { "epoch": 1.2718446601941746, "grad_norm": 28.512968063354492, "learning_rate": 3.037436407034005e-06, "loss": 0.719, "step": 1671 }, { "epoch": 1.272606129830573, "grad_norm": 29.759382247924805, "learning_rate": 3.031833775096251e-06, "loss": 0.6268, "step": 1672 }, { "epoch": 1.2733675994669713, "grad_norm": 39.26443099975586, "learning_rate": 3.026234065935222e-06, "loss": 0.8616, "step": 1673 }, { "epoch": 1.2741290691033695, "grad_norm": 38.28977584838867, "learning_rate": 3.0206372878666267e-06, "loss": 0.6686, "step": 1674 }, { "epoch": 1.2748905387397678, "grad_norm": 28.3647403717041, "learning_rate": 3.015043449201826e-06, "loss": 0.6051, "step": 1675 }, { "epoch": 1.275652008376166, "grad_norm": 20.786691665649414, "learning_rate": 3.0094525582478107e-06, "loss": 0.5821, "step": 1676 }, { "epoch": 1.2764134780125642, "grad_norm": 31.8016414642334, "learning_rate": 3.0038646233071937e-06, "loss": 0.8078, "step": 1677 }, { "epoch": 1.2771749476489624, "grad_norm": 59.46492385864258, "learning_rate": 2.998279652678203e-06, "loss": 0.7843, "step": 1678 }, { "epoch": 1.2779364172853607, "grad_norm": 29.785343170166016, "learning_rate": 2.9926976546546594e-06, "loss": 0.7753, "step": 1679 }, { "epoch": 1.278697886921759, "grad_norm": 31.528043746948242, "learning_rate": 2.987118637525973e-06, "loss": 0.7706, "step": 1680 }, { "epoch": 1.2794593565581573, "grad_norm": 26.38228416442871, "learning_rate": 2.9815426095771234e-06, "loss": 0.6656, "step": 1681 }, { "epoch": 1.2802208261945556, "grad_norm": 25.82967185974121, "learning_rate": 2.9759695790886554e-06, "loss": 0.5333, "step": 1682 }, { "epoch": 1.2809822958309538, "grad_norm": 26.96601676940918, "learning_rate": 2.97039955433666e-06, "loss": 0.4786, "step": 1683 }, { "epoch": 1.281743765467352, "grad_norm": 32.65317153930664, "learning_rate": 2.964832543592766e-06, "loss": 0.5243, "step": 1684 }, { "epoch": 1.2825052351037503, "grad_norm": 30.47889518737793, "learning_rate": 2.959268555124122e-06, "loss": 0.7084, "step": 1685 }, { "epoch": 1.2832667047401485, "grad_norm": 24.077775955200195, "learning_rate": 2.953707597193396e-06, "loss": 0.527, "step": 1686 }, { "epoch": 1.2840281743765467, "grad_norm": 26.070640563964844, "learning_rate": 2.9481496780587504e-06, "loss": 0.6204, "step": 1687 }, { "epoch": 1.284789644012945, "grad_norm": 24.488290786743164, "learning_rate": 2.9425948059738347e-06, "loss": 0.6239, "step": 1688 }, { "epoch": 1.2855511136493432, "grad_norm": 26.81791877746582, "learning_rate": 2.937042989187774e-06, "loss": 0.5609, "step": 1689 }, { "epoch": 1.2863125832857416, "grad_norm": 27.45941162109375, "learning_rate": 2.9314942359451598e-06, "loss": 0.7064, "step": 1690 }, { "epoch": 1.2870740529221396, "grad_norm": 21.853904724121094, "learning_rate": 2.9259485544860313e-06, "loss": 0.45, "step": 1691 }, { "epoch": 1.287835522558538, "grad_norm": 29.07480812072754, "learning_rate": 2.9204059530458617e-06, "loss": 0.7462, "step": 1692 }, { "epoch": 1.2885969921949363, "grad_norm": 27.80030059814453, "learning_rate": 2.914866439855558e-06, "loss": 0.8176, "step": 1693 }, { "epoch": 1.2893584618313345, "grad_norm": 31.991365432739258, "learning_rate": 2.9093300231414384e-06, "loss": 0.5985, "step": 1694 }, { "epoch": 1.2901199314677327, "grad_norm": 27.99630355834961, "learning_rate": 2.9037967111252204e-06, "loss": 0.541, "step": 1695 }, { "epoch": 1.290881401104131, "grad_norm": 33.25284957885742, "learning_rate": 2.898266512024013e-06, "loss": 0.9568, "step": 1696 }, { "epoch": 1.2916428707405292, "grad_norm": 27.58580780029297, "learning_rate": 2.892739434050302e-06, "loss": 0.6486, "step": 1697 }, { "epoch": 1.2924043403769274, "grad_norm": 40.91666793823242, "learning_rate": 2.887215485411938e-06, "loss": 0.9498, "step": 1698 }, { "epoch": 1.2931658100133256, "grad_norm": 37.775630950927734, "learning_rate": 2.8816946743121255e-06, "loss": 0.9751, "step": 1699 }, { "epoch": 1.2939272796497239, "grad_norm": 26.13059425354004, "learning_rate": 2.876177008949408e-06, "loss": 0.7536, "step": 1700 }, { "epoch": 1.2946887492861223, "grad_norm": 30.826940536499023, "learning_rate": 2.870662497517659e-06, "loss": 0.8132, "step": 1701 }, { "epoch": 1.2954502189225205, "grad_norm": 29.84871482849121, "learning_rate": 2.865151148206067e-06, "loss": 0.6199, "step": 1702 }, { "epoch": 1.2962116885589188, "grad_norm": 27.338834762573242, "learning_rate": 2.859642969199127e-06, "loss": 0.6204, "step": 1703 }, { "epoch": 1.296973158195317, "grad_norm": 26.12261962890625, "learning_rate": 2.854137968676622e-06, "loss": 0.6051, "step": 1704 }, { "epoch": 1.2977346278317152, "grad_norm": 33.6061897277832, "learning_rate": 2.84863615481362e-06, "loss": 0.7589, "step": 1705 }, { "epoch": 1.2984960974681135, "grad_norm": 28.311887741088867, "learning_rate": 2.8431375357804517e-06, "loss": 0.7503, "step": 1706 }, { "epoch": 1.2992575671045117, "grad_norm": 28.240816116333008, "learning_rate": 2.8376421197427074e-06, "loss": 0.6389, "step": 1707 }, { "epoch": 1.30001903674091, "grad_norm": 30.9921817779541, "learning_rate": 2.8321499148612186e-06, "loss": 0.6533, "step": 1708 }, { "epoch": 1.3007805063773081, "grad_norm": 34.143287658691406, "learning_rate": 2.8266609292920477e-06, "loss": 0.7499, "step": 1709 }, { "epoch": 1.3015419760137066, "grad_norm": 24.53726577758789, "learning_rate": 2.8211751711864794e-06, "loss": 0.6467, "step": 1710 }, { "epoch": 1.3023034456501046, "grad_norm": 27.959331512451172, "learning_rate": 2.815692648691002e-06, "loss": 0.8354, "step": 1711 }, { "epoch": 1.303064915286503, "grad_norm": 29.361528396606445, "learning_rate": 2.8102133699473012e-06, "loss": 0.7026, "step": 1712 }, { "epoch": 1.3038263849229013, "grad_norm": 24.857891082763672, "learning_rate": 2.804737343092244e-06, "loss": 0.6436, "step": 1713 }, { "epoch": 1.3045878545592995, "grad_norm": 36.500160217285156, "learning_rate": 2.7992645762578697e-06, "loss": 0.8103, "step": 1714 }, { "epoch": 1.3053493241956977, "grad_norm": 22.79261016845703, "learning_rate": 2.793795077571375e-06, "loss": 0.7247, "step": 1715 }, { "epoch": 1.306110793832096, "grad_norm": 26.849435806274414, "learning_rate": 2.788328855155106e-06, "loss": 0.6309, "step": 1716 }, { "epoch": 1.3068722634684942, "grad_norm": 22.991931915283203, "learning_rate": 2.7828659171265387e-06, "loss": 0.6879, "step": 1717 }, { "epoch": 1.3076337331048924, "grad_norm": 33.15441131591797, "learning_rate": 2.777406271598276e-06, "loss": 0.8451, "step": 1718 }, { "epoch": 1.3083952027412906, "grad_norm": 25.909330368041992, "learning_rate": 2.771949926678031e-06, "loss": 0.5737, "step": 1719 }, { "epoch": 1.3091566723776888, "grad_norm": 28.60917091369629, "learning_rate": 2.7664968904686126e-06, "loss": 0.779, "step": 1720 }, { "epoch": 1.3099181420140873, "grad_norm": 27.558080673217773, "learning_rate": 2.7610471710679177e-06, "loss": 0.732, "step": 1721 }, { "epoch": 1.3106796116504853, "grad_norm": 24.676116943359375, "learning_rate": 2.7556007765689183e-06, "loss": 0.6093, "step": 1722 }, { "epoch": 1.3114410812868837, "grad_norm": 30.48960304260254, "learning_rate": 2.750157715059649e-06, "loss": 0.6829, "step": 1723 }, { "epoch": 1.312202550923282, "grad_norm": 27.942277908325195, "learning_rate": 2.7447179946231926e-06, "loss": 0.6584, "step": 1724 }, { "epoch": 1.3129640205596802, "grad_norm": 30.003002166748047, "learning_rate": 2.7392816233376723e-06, "loss": 0.6847, "step": 1725 }, { "epoch": 1.3137254901960784, "grad_norm": 25.514511108398438, "learning_rate": 2.733848609276238e-06, "loss": 0.6992, "step": 1726 }, { "epoch": 1.3144869598324767, "grad_norm": 26.44031524658203, "learning_rate": 2.7284189605070523e-06, "loss": 0.6089, "step": 1727 }, { "epoch": 1.3152484294688749, "grad_norm": 30.96246337890625, "learning_rate": 2.722992685093282e-06, "loss": 0.6525, "step": 1728 }, { "epoch": 1.316009899105273, "grad_norm": 34.42742156982422, "learning_rate": 2.7175697910930832e-06, "loss": 0.8125, "step": 1729 }, { "epoch": 1.3167713687416716, "grad_norm": 24.77292823791504, "learning_rate": 2.7121502865595905e-06, "loss": 0.6029, "step": 1730 }, { "epoch": 1.3175328383780696, "grad_norm": 29.337425231933594, "learning_rate": 2.706734179540907e-06, "loss": 0.714, "step": 1731 }, { "epoch": 1.318294308014468, "grad_norm": 32.2049560546875, "learning_rate": 2.7013214780800875e-06, "loss": 0.8105, "step": 1732 }, { "epoch": 1.3190557776508662, "grad_norm": 31.802593231201172, "learning_rate": 2.695912190215131e-06, "loss": 0.6363, "step": 1733 }, { "epoch": 1.3198172472872645, "grad_norm": 27.509658813476562, "learning_rate": 2.6905063239789673e-06, "loss": 0.5398, "step": 1734 }, { "epoch": 1.3205787169236627, "grad_norm": 30.282777786254883, "learning_rate": 2.6851038873994452e-06, "loss": 0.854, "step": 1735 }, { "epoch": 1.321340186560061, "grad_norm": 26.111318588256836, "learning_rate": 2.6797048884993165e-06, "loss": 0.6994, "step": 1736 }, { "epoch": 1.3221016561964591, "grad_norm": 26.988555908203125, "learning_rate": 2.6743093352962392e-06, "loss": 0.829, "step": 1737 }, { "epoch": 1.3228631258328574, "grad_norm": 26.739686965942383, "learning_rate": 2.6689172358027394e-06, "loss": 0.4448, "step": 1738 }, { "epoch": 1.3236245954692556, "grad_norm": 28.498573303222656, "learning_rate": 2.6635285980262243e-06, "loss": 0.687, "step": 1739 }, { "epoch": 1.3243860651056538, "grad_norm": 30.018733978271484, "learning_rate": 2.658143429968957e-06, "loss": 0.5537, "step": 1740 }, { "epoch": 1.3251475347420523, "grad_norm": 26.414478302001953, "learning_rate": 2.652761739628048e-06, "loss": 0.6008, "step": 1741 }, { "epoch": 1.3259090043784503, "grad_norm": 26.425273895263672, "learning_rate": 2.647383534995446e-06, "loss": 0.6315, "step": 1742 }, { "epoch": 1.3266704740148487, "grad_norm": 25.95477867126465, "learning_rate": 2.6420088240579185e-06, "loss": 0.6062, "step": 1743 }, { "epoch": 1.327431943651247, "grad_norm": 24.891244888305664, "learning_rate": 2.636637614797049e-06, "loss": 0.5731, "step": 1744 }, { "epoch": 1.3281934132876452, "grad_norm": 27.093591690063477, "learning_rate": 2.631269915189219e-06, "loss": 0.663, "step": 1745 }, { "epoch": 1.3289548829240434, "grad_norm": 28.390575408935547, "learning_rate": 2.6259057332055987e-06, "loss": 0.7888, "step": 1746 }, { "epoch": 1.3297163525604416, "grad_norm": 26.620412826538086, "learning_rate": 2.620545076812132e-06, "loss": 0.5802, "step": 1747 }, { "epoch": 1.3304778221968399, "grad_norm": 31.760650634765625, "learning_rate": 2.6151879539695357e-06, "loss": 0.7689, "step": 1748 }, { "epoch": 1.331239291833238, "grad_norm": 25.756576538085938, "learning_rate": 2.6098343726332665e-06, "loss": 0.6005, "step": 1749 }, { "epoch": 1.3320007614696365, "grad_norm": 28.34639549255371, "learning_rate": 2.6044843407535313e-06, "loss": 0.7928, "step": 1750 }, { "epoch": 1.3327622311060345, "grad_norm": 36.99173355102539, "learning_rate": 2.599137866275261e-06, "loss": 0.8049, "step": 1751 }, { "epoch": 1.333523700742433, "grad_norm": 30.13863182067871, "learning_rate": 2.5937949571381103e-06, "loss": 0.6576, "step": 1752 }, { "epoch": 1.3342851703788312, "grad_norm": 28.823902130126953, "learning_rate": 2.58845562127643e-06, "loss": 0.5499, "step": 1753 }, { "epoch": 1.3350466400152294, "grad_norm": 29.00064468383789, "learning_rate": 2.58311986661927e-06, "loss": 0.779, "step": 1754 }, { "epoch": 1.3358081096516277, "grad_norm": 23.507123947143555, "learning_rate": 2.5777877010903605e-06, "loss": 0.636, "step": 1755 }, { "epoch": 1.3365695792880259, "grad_norm": 29.09501838684082, "learning_rate": 2.572459132608108e-06, "loss": 0.6588, "step": 1756 }, { "epoch": 1.3373310489244241, "grad_norm": 29.63762855529785, "learning_rate": 2.5671341690855665e-06, "loss": 0.6565, "step": 1757 }, { "epoch": 1.3380925185608223, "grad_norm": 37.66661834716797, "learning_rate": 2.561812818430442e-06, "loss": 1.0703, "step": 1758 }, { "epoch": 1.3388539881972206, "grad_norm": 28.495466232299805, "learning_rate": 2.5564950885450763e-06, "loss": 0.4935, "step": 1759 }, { "epoch": 1.3396154578336188, "grad_norm": 21.38892936706543, "learning_rate": 2.5511809873264325e-06, "loss": 0.5735, "step": 1760 }, { "epoch": 1.3403769274700172, "grad_norm": 29.09061050415039, "learning_rate": 2.5458705226660864e-06, "loss": 0.7607, "step": 1761 }, { "epoch": 1.3411383971064152, "grad_norm": 28.077455520629883, "learning_rate": 2.5405637024502096e-06, "loss": 0.732, "step": 1762 }, { "epoch": 1.3418998667428137, "grad_norm": 29.965654373168945, "learning_rate": 2.5352605345595714e-06, "loss": 0.642, "step": 1763 }, { "epoch": 1.342661336379212, "grad_norm": 31.610265731811523, "learning_rate": 2.529961026869504e-06, "loss": 0.8096, "step": 1764 }, { "epoch": 1.3434228060156101, "grad_norm": 30.006776809692383, "learning_rate": 2.524665187249913e-06, "loss": 0.7359, "step": 1765 }, { "epoch": 1.3441842756520084, "grad_norm": 30.135059356689453, "learning_rate": 2.5193730235652527e-06, "loss": 0.8384, "step": 1766 }, { "epoch": 1.3449457452884066, "grad_norm": 29.22550392150879, "learning_rate": 2.5140845436745255e-06, "loss": 0.6449, "step": 1767 }, { "epoch": 1.3457072149248048, "grad_norm": 27.91558074951172, "learning_rate": 2.5087997554312525e-06, "loss": 0.8288, "step": 1768 }, { "epoch": 1.346468684561203, "grad_norm": 27.852632522583008, "learning_rate": 2.5035186666834798e-06, "loss": 0.7255, "step": 1769 }, { "epoch": 1.3472301541976015, "grad_norm": 27.12213134765625, "learning_rate": 2.4982412852737575e-06, "loss": 0.6744, "step": 1770 }, { "epoch": 1.3479916238339995, "grad_norm": 34.69364929199219, "learning_rate": 2.492967619039135e-06, "loss": 0.5945, "step": 1771 }, { "epoch": 1.348753093470398, "grad_norm": 28.04036521911621, "learning_rate": 2.487697675811135e-06, "loss": 0.8408, "step": 1772 }, { "epoch": 1.3495145631067962, "grad_norm": 24.792219161987305, "learning_rate": 2.4824314634157564e-06, "loss": 0.6894, "step": 1773 }, { "epoch": 1.3502760327431944, "grad_norm": 33.18821334838867, "learning_rate": 2.4771689896734647e-06, "loss": 0.883, "step": 1774 }, { "epoch": 1.3510375023795926, "grad_norm": 26.50601577758789, "learning_rate": 2.47191026239916e-06, "loss": 0.6698, "step": 1775 }, { "epoch": 1.3517989720159909, "grad_norm": 30.306671142578125, "learning_rate": 2.466655289402189e-06, "loss": 0.5945, "step": 1776 }, { "epoch": 1.352560441652389, "grad_norm": 30.933183670043945, "learning_rate": 2.4614040784863164e-06, "loss": 0.6771, "step": 1777 }, { "epoch": 1.3533219112887873, "grad_norm": 56.53095245361328, "learning_rate": 2.4561566374497296e-06, "loss": 0.7545, "step": 1778 }, { "epoch": 1.3540833809251855, "grad_norm": 32.13953399658203, "learning_rate": 2.4509129740850065e-06, "loss": 0.5236, "step": 1779 }, { "epoch": 1.3548448505615838, "grad_norm": 36.229366302490234, "learning_rate": 2.4456730961791207e-06, "loss": 0.4947, "step": 1780 }, { "epoch": 1.3556063201979822, "grad_norm": 40.86501693725586, "learning_rate": 2.4404370115134236e-06, "loss": 0.8211, "step": 1781 }, { "epoch": 1.3563677898343802, "grad_norm": 35.70888900756836, "learning_rate": 2.4352047278636376e-06, "loss": 0.7389, "step": 1782 }, { "epoch": 1.3571292594707787, "grad_norm": 30.2214298248291, "learning_rate": 2.4299762529998316e-06, "loss": 0.8245, "step": 1783 }, { "epoch": 1.357890729107177, "grad_norm": 30.86892318725586, "learning_rate": 2.4247515946864232e-06, "loss": 0.7508, "step": 1784 }, { "epoch": 1.3586521987435751, "grad_norm": 32.86927032470703, "learning_rate": 2.419530760682166e-06, "loss": 0.9101, "step": 1785 }, { "epoch": 1.3594136683799734, "grad_norm": 30.969940185546875, "learning_rate": 2.414313758740131e-06, "loss": 0.7277, "step": 1786 }, { "epoch": 1.3601751380163716, "grad_norm": 23.66486930847168, "learning_rate": 2.409100596607695e-06, "loss": 0.5382, "step": 1787 }, { "epoch": 1.3609366076527698, "grad_norm": 28.6279296875, "learning_rate": 2.4038912820265353e-06, "loss": 0.6363, "step": 1788 }, { "epoch": 1.361698077289168, "grad_norm": 25.915483474731445, "learning_rate": 2.39868582273262e-06, "loss": 0.609, "step": 1789 }, { "epoch": 1.3624595469255665, "grad_norm": 22.417701721191406, "learning_rate": 2.3934842264561876e-06, "loss": 0.6348, "step": 1790 }, { "epoch": 1.3632210165619645, "grad_norm": 29.688234329223633, "learning_rate": 2.3882865009217377e-06, "loss": 0.6589, "step": 1791 }, { "epoch": 1.363982486198363, "grad_norm": 25.72180938720703, "learning_rate": 2.383092653848023e-06, "loss": 0.5005, "step": 1792 }, { "epoch": 1.3647439558347612, "grad_norm": 28.812807083129883, "learning_rate": 2.377902692948045e-06, "loss": 0.7101, "step": 1793 }, { "epoch": 1.3655054254711594, "grad_norm": 32.06017303466797, "learning_rate": 2.3727166259290213e-06, "loss": 0.6384, "step": 1794 }, { "epoch": 1.3662668951075576, "grad_norm": 23.132009506225586, "learning_rate": 2.367534460492393e-06, "loss": 0.5962, "step": 1795 }, { "epoch": 1.3670283647439558, "grad_norm": 24.653589248657227, "learning_rate": 2.3623562043338093e-06, "loss": 0.7476, "step": 1796 }, { "epoch": 1.367789834380354, "grad_norm": 28.437185287475586, "learning_rate": 2.3571818651431137e-06, "loss": 0.7654, "step": 1797 }, { "epoch": 1.3685513040167523, "grad_norm": 25.68982696533203, "learning_rate": 2.3520114506043266e-06, "loss": 0.6335, "step": 1798 }, { "epoch": 1.3693127736531505, "grad_norm": 36.46854782104492, "learning_rate": 2.346844968395645e-06, "loss": 0.6833, "step": 1799 }, { "epoch": 1.3700742432895487, "grad_norm": 34.33137512207031, "learning_rate": 2.3416824261894287e-06, "loss": 0.7802, "step": 1800 }, { "epoch": 1.3708357129259472, "grad_norm": 40.84291076660156, "learning_rate": 2.336523831652185e-06, "loss": 0.7067, "step": 1801 }, { "epoch": 1.3715971825623452, "grad_norm": 39.46131134033203, "learning_rate": 2.3313691924445524e-06, "loss": 1.0324, "step": 1802 }, { "epoch": 1.3723586521987436, "grad_norm": 36.6191291809082, "learning_rate": 2.326218516221302e-06, "loss": 0.7747, "step": 1803 }, { "epoch": 1.3731201218351419, "grad_norm": 38.23823165893555, "learning_rate": 2.3210718106313213e-06, "loss": 0.9964, "step": 1804 }, { "epoch": 1.37388159147154, "grad_norm": 38.139869689941406, "learning_rate": 2.315929083317599e-06, "loss": 0.81, "step": 1805 }, { "epoch": 1.3746430611079383, "grad_norm": 34.04991149902344, "learning_rate": 2.31079034191721e-06, "loss": 0.7102, "step": 1806 }, { "epoch": 1.3754045307443366, "grad_norm": 27.099498748779297, "learning_rate": 2.3056555940613212e-06, "loss": 0.6629, "step": 1807 }, { "epoch": 1.3761660003807348, "grad_norm": 30.62109375, "learning_rate": 2.300524847375163e-06, "loss": 0.6754, "step": 1808 }, { "epoch": 1.376927470017133, "grad_norm": 27.176870346069336, "learning_rate": 2.2953981094780204e-06, "loss": 0.7115, "step": 1809 }, { "epoch": 1.3776889396535315, "grad_norm": 30.748756408691406, "learning_rate": 2.2902753879832296e-06, "loss": 0.6005, "step": 1810 }, { "epoch": 1.3784504092899295, "grad_norm": 26.850910186767578, "learning_rate": 2.2851566904981655e-06, "loss": 0.6969, "step": 1811 }, { "epoch": 1.379211878926328, "grad_norm": 23.646045684814453, "learning_rate": 2.280042024624223e-06, "loss": 0.5814, "step": 1812 }, { "epoch": 1.3799733485627261, "grad_norm": 36.72319412231445, "learning_rate": 2.274931397956807e-06, "loss": 0.7331, "step": 1813 }, { "epoch": 1.3807348181991244, "grad_norm": 26.8786563873291, "learning_rate": 2.2698248180853287e-06, "loss": 0.6315, "step": 1814 }, { "epoch": 1.3814962878355226, "grad_norm": 30.796337127685547, "learning_rate": 2.2647222925931912e-06, "loss": 0.7601, "step": 1815 }, { "epoch": 1.3822577574719208, "grad_norm": 26.428974151611328, "learning_rate": 2.2596238290577743e-06, "loss": 0.5928, "step": 1816 }, { "epoch": 1.383019227108319, "grad_norm": 36.869503021240234, "learning_rate": 2.25452943505042e-06, "loss": 0.8022, "step": 1817 }, { "epoch": 1.3837806967447173, "grad_norm": 25.970426559448242, "learning_rate": 2.249439118136439e-06, "loss": 0.6983, "step": 1818 }, { "epoch": 1.3845421663811155, "grad_norm": 33.53501510620117, "learning_rate": 2.2443528858750775e-06, "loss": 0.8529, "step": 1819 }, { "epoch": 1.3853036360175137, "grad_norm": 38.37016296386719, "learning_rate": 2.239270745819522e-06, "loss": 0.7786, "step": 1820 }, { "epoch": 1.3860651056539122, "grad_norm": 27.427955627441406, "learning_rate": 2.2341927055168746e-06, "loss": 0.6804, "step": 1821 }, { "epoch": 1.3868265752903102, "grad_norm": 32.579769134521484, "learning_rate": 2.2291187725081582e-06, "loss": 0.8268, "step": 1822 }, { "epoch": 1.3875880449267086, "grad_norm": 24.318727493286133, "learning_rate": 2.22404895432829e-06, "loss": 0.6072, "step": 1823 }, { "epoch": 1.3883495145631068, "grad_norm": 23.48029899597168, "learning_rate": 2.218983258506081e-06, "loss": 0.5016, "step": 1824 }, { "epoch": 1.389110984199505, "grad_norm": 46.418846130371094, "learning_rate": 2.213921692564212e-06, "loss": 0.8553, "step": 1825 }, { "epoch": 1.3898724538359033, "grad_norm": 27.075538635253906, "learning_rate": 2.2088642640192426e-06, "loss": 0.5535, "step": 1826 }, { "epoch": 1.3906339234723015, "grad_norm": 33.25198745727539, "learning_rate": 2.2038109803815816e-06, "loss": 0.8656, "step": 1827 }, { "epoch": 1.3913953931086998, "grad_norm": 35.50556564331055, "learning_rate": 2.1987618491554784e-06, "loss": 0.8391, "step": 1828 }, { "epoch": 1.392156862745098, "grad_norm": 29.472679138183594, "learning_rate": 2.1937168778390268e-06, "loss": 0.7238, "step": 1829 }, { "epoch": 1.3929183323814962, "grad_norm": 29.225574493408203, "learning_rate": 2.1886760739241335e-06, "loss": 0.7497, "step": 1830 }, { "epoch": 1.3936798020178944, "grad_norm": 26.468120574951172, "learning_rate": 2.1836394448965243e-06, "loss": 0.5661, "step": 1831 }, { "epoch": 1.3944412716542929, "grad_norm": 29.12610626220703, "learning_rate": 2.178606998235713e-06, "loss": 0.5745, "step": 1832 }, { "epoch": 1.395202741290691, "grad_norm": 24.718732833862305, "learning_rate": 2.173578741415018e-06, "loss": 0.6124, "step": 1833 }, { "epoch": 1.3959642109270893, "grad_norm": 31.89914321899414, "learning_rate": 2.168554681901525e-06, "loss": 0.8232, "step": 1834 }, { "epoch": 1.3967256805634876, "grad_norm": 38.25407028198242, "learning_rate": 2.163534827156092e-06, "loss": 0.7579, "step": 1835 }, { "epoch": 1.3974871501998858, "grad_norm": 29.516969680786133, "learning_rate": 2.1585191846333247e-06, "loss": 0.7496, "step": 1836 }, { "epoch": 1.398248619836284, "grad_norm": 28.61370086669922, "learning_rate": 2.153507761781585e-06, "loss": 0.847, "step": 1837 }, { "epoch": 1.3990100894726822, "grad_norm": 25.529470443725586, "learning_rate": 2.148500566042962e-06, "loss": 0.5238, "step": 1838 }, { "epoch": 1.3997715591090805, "grad_norm": 27.421363830566406, "learning_rate": 2.1434976048532685e-06, "loss": 0.6547, "step": 1839 }, { "epoch": 1.4005330287454787, "grad_norm": 42.281471252441406, "learning_rate": 2.1384988856420243e-06, "loss": 0.8199, "step": 1840 }, { "epoch": 1.4012944983818771, "grad_norm": 24.29891014099121, "learning_rate": 2.1335044158324593e-06, "loss": 0.5808, "step": 1841 }, { "epoch": 1.4020559680182751, "grad_norm": 28.176315307617188, "learning_rate": 2.1285142028414875e-06, "loss": 0.564, "step": 1842 }, { "epoch": 1.4028174376546736, "grad_norm": 31.240074157714844, "learning_rate": 2.123528254079696e-06, "loss": 0.6914, "step": 1843 }, { "epoch": 1.4035789072910718, "grad_norm": 30.84389305114746, "learning_rate": 2.1185465769513507e-06, "loss": 0.9383, "step": 1844 }, { "epoch": 1.40434037692747, "grad_norm": 38.649051666259766, "learning_rate": 2.1135691788543676e-06, "loss": 0.6248, "step": 1845 }, { "epoch": 1.4051018465638683, "grad_norm": 27.57621955871582, "learning_rate": 2.1085960671803094e-06, "loss": 0.5741, "step": 1846 }, { "epoch": 1.4058633162002665, "grad_norm": 24.17763328552246, "learning_rate": 2.1036272493143684e-06, "loss": 0.5881, "step": 1847 }, { "epoch": 1.4066247858366647, "grad_norm": 39.37004852294922, "learning_rate": 2.0986627326353703e-06, "loss": 0.7304, "step": 1848 }, { "epoch": 1.407386255473063, "grad_norm": 32.73893356323242, "learning_rate": 2.093702524515746e-06, "loss": 0.7724, "step": 1849 }, { "epoch": 1.4081477251094612, "grad_norm": 28.6301212310791, "learning_rate": 2.0887466323215326e-06, "loss": 0.6428, "step": 1850 }, { "epoch": 1.4089091947458594, "grad_norm": 26.686613082885742, "learning_rate": 2.0837950634123503e-06, "loss": 0.6267, "step": 1851 }, { "epoch": 1.4096706643822579, "grad_norm": 40.792476654052734, "learning_rate": 2.0788478251414092e-06, "loss": 0.8664, "step": 1852 }, { "epoch": 1.410432134018656, "grad_norm": 30.36296272277832, "learning_rate": 2.073904924855482e-06, "loss": 0.7543, "step": 1853 }, { "epoch": 1.4111936036550543, "grad_norm": 22.311126708984375, "learning_rate": 2.068966369894901e-06, "loss": 0.4556, "step": 1854 }, { "epoch": 1.4119550732914525, "grad_norm": 34.04696273803711, "learning_rate": 2.0640321675935456e-06, "loss": 0.8219, "step": 1855 }, { "epoch": 1.4127165429278508, "grad_norm": 27.769725799560547, "learning_rate": 2.059102325278831e-06, "loss": 0.7987, "step": 1856 }, { "epoch": 1.413478012564249, "grad_norm": 25.73517608642578, "learning_rate": 2.054176850271698e-06, "loss": 0.7236, "step": 1857 }, { "epoch": 1.4142394822006472, "grad_norm": 32.25333786010742, "learning_rate": 2.0492557498866016e-06, "loss": 0.7241, "step": 1858 }, { "epoch": 1.4150009518370454, "grad_norm": 30.307878494262695, "learning_rate": 2.0443390314315013e-06, "loss": 0.7375, "step": 1859 }, { "epoch": 1.4157624214734437, "grad_norm": 30.474096298217773, "learning_rate": 2.039426702207848e-06, "loss": 0.631, "step": 1860 }, { "epoch": 1.4165238911098421, "grad_norm": 25.01543426513672, "learning_rate": 2.0345187695105774e-06, "loss": 0.584, "step": 1861 }, { "epoch": 1.4172853607462401, "grad_norm": 33.293766021728516, "learning_rate": 2.029615240628088e-06, "loss": 0.7583, "step": 1862 }, { "epoch": 1.4180468303826386, "grad_norm": 30.636577606201172, "learning_rate": 2.0247161228422503e-06, "loss": 0.7273, "step": 1863 }, { "epoch": 1.4188083000190368, "grad_norm": 28.93274688720703, "learning_rate": 2.019821423428376e-06, "loss": 0.5972, "step": 1864 }, { "epoch": 1.419569769655435, "grad_norm": 28.450654983520508, "learning_rate": 2.0149311496552183e-06, "loss": 0.6593, "step": 1865 }, { "epoch": 1.4203312392918332, "grad_norm": 25.276092529296875, "learning_rate": 2.0100453087849576e-06, "loss": 0.4372, "step": 1866 }, { "epoch": 1.4210927089282315, "grad_norm": 26.568655014038086, "learning_rate": 2.005163908073191e-06, "loss": 0.7544, "step": 1867 }, { "epoch": 1.4218541785646297, "grad_norm": 35.838809967041016, "learning_rate": 2.000286954768923e-06, "loss": 0.8063, "step": 1868 }, { "epoch": 1.422615648201028, "grad_norm": 25.219032287597656, "learning_rate": 1.9954144561145526e-06, "loss": 0.5946, "step": 1869 }, { "epoch": 1.4233771178374262, "grad_norm": 25.65085220336914, "learning_rate": 1.990546419345864e-06, "loss": 0.4989, "step": 1870 }, { "epoch": 1.4241385874738244, "grad_norm": 22.25298500061035, "learning_rate": 1.985682851692014e-06, "loss": 0.5515, "step": 1871 }, { "epoch": 1.4249000571102228, "grad_norm": 33.739200592041016, "learning_rate": 1.9808237603755252e-06, "loss": 0.7662, "step": 1872 }, { "epoch": 1.425661526746621, "grad_norm": 27.28119468688965, "learning_rate": 1.975969152612271e-06, "loss": 0.8033, "step": 1873 }, { "epoch": 1.4264229963830193, "grad_norm": 27.47292137145996, "learning_rate": 1.971119035611467e-06, "loss": 0.5678, "step": 1874 }, { "epoch": 1.4271844660194175, "grad_norm": 28.936887741088867, "learning_rate": 1.966273416575659e-06, "loss": 0.6715, "step": 1875 }, { "epoch": 1.4279459356558157, "grad_norm": 22.802989959716797, "learning_rate": 1.9614323027007153e-06, "loss": 0.5158, "step": 1876 }, { "epoch": 1.428707405292214, "grad_norm": 21.903430938720703, "learning_rate": 1.956595701175812e-06, "loss": 0.5922, "step": 1877 }, { "epoch": 1.4294688749286122, "grad_norm": 28.148441314697266, "learning_rate": 1.951763619183425e-06, "loss": 0.7152, "step": 1878 }, { "epoch": 1.4302303445650104, "grad_norm": 23.71442222595215, "learning_rate": 1.946936063899319e-06, "loss": 0.6005, "step": 1879 }, { "epoch": 1.4309918142014086, "grad_norm": 24.66382598876953, "learning_rate": 1.942113042492535e-06, "loss": 0.6093, "step": 1880 }, { "epoch": 1.431753283837807, "grad_norm": 35.267513275146484, "learning_rate": 1.937294562125383e-06, "loss": 0.7775, "step": 1881 }, { "epoch": 1.432514753474205, "grad_norm": 25.363155364990234, "learning_rate": 1.9324806299534284e-06, "loss": 0.518, "step": 1882 }, { "epoch": 1.4332762231106035, "grad_norm": 27.412708282470703, "learning_rate": 1.9276712531254816e-06, "loss": 0.755, "step": 1883 }, { "epoch": 1.4340376927470018, "grad_norm": 37.167381286621094, "learning_rate": 1.92286643878359e-06, "loss": 0.8745, "step": 1884 }, { "epoch": 1.4347991623834, "grad_norm": 25.316354751586914, "learning_rate": 1.918066194063024e-06, "loss": 0.7132, "step": 1885 }, { "epoch": 1.4355606320197982, "grad_norm": 31.473587036132812, "learning_rate": 1.913270526092269e-06, "loss": 0.7547, "step": 1886 }, { "epoch": 1.4363221016561964, "grad_norm": 25.11524200439453, "learning_rate": 1.9084794419930134e-06, "loss": 0.7424, "step": 1887 }, { "epoch": 1.4370835712925947, "grad_norm": 23.20271110534668, "learning_rate": 1.9036929488801387e-06, "loss": 0.5298, "step": 1888 }, { "epoch": 1.437845040928993, "grad_norm": 27.01667022705078, "learning_rate": 1.8989110538617083e-06, "loss": 0.6513, "step": 1889 }, { "epoch": 1.4386065105653911, "grad_norm": 27.438274383544922, "learning_rate": 1.8941337640389579e-06, "loss": 0.6097, "step": 1890 }, { "epoch": 1.4393679802017894, "grad_norm": 32.6960334777832, "learning_rate": 1.8893610865062834e-06, "loss": 0.8089, "step": 1891 }, { "epoch": 1.4401294498381878, "grad_norm": 26.43588638305664, "learning_rate": 1.8845930283512316e-06, "loss": 0.6655, "step": 1892 }, { "epoch": 1.440890919474586, "grad_norm": 35.13780975341797, "learning_rate": 1.8798295966544905e-06, "loss": 0.6858, "step": 1893 }, { "epoch": 1.4416523891109843, "grad_norm": 28.761791229248047, "learning_rate": 1.8750707984898769e-06, "loss": 0.5817, "step": 1894 }, { "epoch": 1.4424138587473825, "grad_norm": 29.459430694580078, "learning_rate": 1.870316640924326e-06, "loss": 0.6822, "step": 1895 }, { "epoch": 1.4431753283837807, "grad_norm": 31.26158332824707, "learning_rate": 1.865567131017883e-06, "loss": 0.7195, "step": 1896 }, { "epoch": 1.443936798020179, "grad_norm": 34.26539611816406, "learning_rate": 1.8608222758236898e-06, "loss": 0.6405, "step": 1897 }, { "epoch": 1.4446982676565772, "grad_norm": 25.34882354736328, "learning_rate": 1.8560820823879776e-06, "loss": 0.578, "step": 1898 }, { "epoch": 1.4454597372929754, "grad_norm": 32.722137451171875, "learning_rate": 1.8513465577500527e-06, "loss": 0.6309, "step": 1899 }, { "epoch": 1.4462212069293736, "grad_norm": 23.928403854370117, "learning_rate": 1.84661570894229e-06, "loss": 0.7601, "step": 1900 }, { "epoch": 1.446982676565772, "grad_norm": 31.33304214477539, "learning_rate": 1.8418895429901191e-06, "loss": 0.751, "step": 1901 }, { "epoch": 1.44774414620217, "grad_norm": 29.492887496948242, "learning_rate": 1.8371680669120162e-06, "loss": 0.5471, "step": 1902 }, { "epoch": 1.4485056158385685, "grad_norm": 34.49226379394531, "learning_rate": 1.832451287719494e-06, "loss": 0.9174, "step": 1903 }, { "epoch": 1.4492670854749667, "grad_norm": 38.36891555786133, "learning_rate": 1.8277392124170884e-06, "loss": 0.8829, "step": 1904 }, { "epoch": 1.450028555111365, "grad_norm": 31.16155433654785, "learning_rate": 1.8230318480023512e-06, "loss": 0.6943, "step": 1905 }, { "epoch": 1.4507900247477632, "grad_norm": 85.44965362548828, "learning_rate": 1.8183292014658372e-06, "loss": 0.8847, "step": 1906 }, { "epoch": 1.4515514943841614, "grad_norm": 32.240020751953125, "learning_rate": 1.8136312797910955e-06, "loss": 0.7322, "step": 1907 }, { "epoch": 1.4523129640205596, "grad_norm": 35.49117660522461, "learning_rate": 1.8089380899546599e-06, "loss": 0.8246, "step": 1908 }, { "epoch": 1.4530744336569579, "grad_norm": 32.99125671386719, "learning_rate": 1.804249638926035e-06, "loss": 0.7992, "step": 1909 }, { "epoch": 1.453835903293356, "grad_norm": 29.68242645263672, "learning_rate": 1.7995659336676907e-06, "loss": 0.7837, "step": 1910 }, { "epoch": 1.4545973729297543, "grad_norm": 28.061498641967773, "learning_rate": 1.7948869811350472e-06, "loss": 0.7148, "step": 1911 }, { "epoch": 1.4553588425661528, "grad_norm": 29.588260650634766, "learning_rate": 1.7902127882764676e-06, "loss": 0.6517, "step": 1912 }, { "epoch": 1.456120312202551, "grad_norm": 30.845842361450195, "learning_rate": 1.785543362033247e-06, "loss": 0.6508, "step": 1913 }, { "epoch": 1.4568817818389492, "grad_norm": 37.057682037353516, "learning_rate": 1.7808787093396013e-06, "loss": 0.6909, "step": 1914 }, { "epoch": 1.4576432514753475, "grad_norm": 28.16147804260254, "learning_rate": 1.7762188371226585e-06, "loss": 0.6442, "step": 1915 }, { "epoch": 1.4584047211117457, "grad_norm": 29.85873031616211, "learning_rate": 1.771563752302447e-06, "loss": 0.6137, "step": 1916 }, { "epoch": 1.459166190748144, "grad_norm": 33.01405334472656, "learning_rate": 1.7669134617918847e-06, "loss": 0.6233, "step": 1917 }, { "epoch": 1.4599276603845421, "grad_norm": 28.902408599853516, "learning_rate": 1.7622679724967729e-06, "loss": 0.7603, "step": 1918 }, { "epoch": 1.4606891300209404, "grad_norm": 21.834259033203125, "learning_rate": 1.7576272913157789e-06, "loss": 0.5494, "step": 1919 }, { "epoch": 1.4614505996573386, "grad_norm": 29.060260772705078, "learning_rate": 1.7529914251404322e-06, "loss": 0.753, "step": 1920 }, { "epoch": 1.462212069293737, "grad_norm": 39.84613037109375, "learning_rate": 1.7483603808551131e-06, "loss": 0.8935, "step": 1921 }, { "epoch": 1.462973538930135, "grad_norm": 36.62184524536133, "learning_rate": 1.7437341653370377e-06, "loss": 0.8394, "step": 1922 }, { "epoch": 1.4637350085665335, "grad_norm": 31.159271240234375, "learning_rate": 1.7391127854562544e-06, "loss": 0.7482, "step": 1923 }, { "epoch": 1.4644964782029317, "grad_norm": 38.46657180786133, "learning_rate": 1.7344962480756284e-06, "loss": 0.8478, "step": 1924 }, { "epoch": 1.46525794783933, "grad_norm": 28.594711303710938, "learning_rate": 1.729884560050838e-06, "loss": 0.7019, "step": 1925 }, { "epoch": 1.4660194174757282, "grad_norm": 28.661853790283203, "learning_rate": 1.7252777282303535e-06, "loss": 0.6434, "step": 1926 }, { "epoch": 1.4667808871121264, "grad_norm": 22.471424102783203, "learning_rate": 1.7206757594554375e-06, "loss": 0.5731, "step": 1927 }, { "epoch": 1.4675423567485246, "grad_norm": 31.315664291381836, "learning_rate": 1.7160786605601305e-06, "loss": 0.7715, "step": 1928 }, { "epoch": 1.4683038263849228, "grad_norm": 22.22355079650879, "learning_rate": 1.7114864383712403e-06, "loss": 0.4805, "step": 1929 }, { "epoch": 1.469065296021321, "grad_norm": 24.957242965698242, "learning_rate": 1.7068990997083346e-06, "loss": 0.5891, "step": 1930 }, { "epoch": 1.4698267656577193, "grad_norm": 29.125608444213867, "learning_rate": 1.7023166513837268e-06, "loss": 0.6257, "step": 1931 }, { "epoch": 1.4705882352941178, "grad_norm": 30.713321685791016, "learning_rate": 1.697739100202469e-06, "loss": 0.7315, "step": 1932 }, { "epoch": 1.471349704930516, "grad_norm": 27.89665412902832, "learning_rate": 1.6931664529623404e-06, "loss": 0.7149, "step": 1933 }, { "epoch": 1.4721111745669142, "grad_norm": 23.54539680480957, "learning_rate": 1.6885987164538382e-06, "loss": 0.4831, "step": 1934 }, { "epoch": 1.4728726442033124, "grad_norm": 31.730077743530273, "learning_rate": 1.684035897460165e-06, "loss": 0.7505, "step": 1935 }, { "epoch": 1.4736341138397107, "grad_norm": 40.70829772949219, "learning_rate": 1.6794780027572282e-06, "loss": 0.9372, "step": 1936 }, { "epoch": 1.4743955834761089, "grad_norm": 27.258739471435547, "learning_rate": 1.6749250391136112e-06, "loss": 0.6753, "step": 1937 }, { "epoch": 1.475157053112507, "grad_norm": 30.2043514251709, "learning_rate": 1.6703770132905834e-06, "loss": 0.8191, "step": 1938 }, { "epoch": 1.4759185227489053, "grad_norm": 32.73941421508789, "learning_rate": 1.6658339320420751e-06, "loss": 0.7147, "step": 1939 }, { "epoch": 1.4766799923853036, "grad_norm": 26.30469512939453, "learning_rate": 1.661295802114683e-06, "loss": 0.6912, "step": 1940 }, { "epoch": 1.477441462021702, "grad_norm": 28.07451057434082, "learning_rate": 1.6567626302476397e-06, "loss": 0.6439, "step": 1941 }, { "epoch": 1.4782029316581, "grad_norm": 25.540842056274414, "learning_rate": 1.6522344231728221e-06, "loss": 0.6091, "step": 1942 }, { "epoch": 1.4789644012944985, "grad_norm": 27.23672866821289, "learning_rate": 1.6477111876147328e-06, "loss": 0.5726, "step": 1943 }, { "epoch": 1.4797258709308967, "grad_norm": 27.432655334472656, "learning_rate": 1.6431929302904914e-06, "loss": 0.6402, "step": 1944 }, { "epoch": 1.480487340567295, "grad_norm": 30.246688842773438, "learning_rate": 1.6386796579098236e-06, "loss": 0.8161, "step": 1945 }, { "epoch": 1.4812488102036931, "grad_norm": 30.486093521118164, "learning_rate": 1.6341713771750523e-06, "loss": 0.6322, "step": 1946 }, { "epoch": 1.4820102798400914, "grad_norm": 36.072975158691406, "learning_rate": 1.6296680947810933e-06, "loss": 0.6327, "step": 1947 }, { "epoch": 1.4827717494764896, "grad_norm": 32.71725845336914, "learning_rate": 1.62516981741543e-06, "loss": 0.6791, "step": 1948 }, { "epoch": 1.4835332191128878, "grad_norm": 33.857574462890625, "learning_rate": 1.6206765517581207e-06, "loss": 0.7739, "step": 1949 }, { "epoch": 1.484294688749286, "grad_norm": 27.191062927246094, "learning_rate": 1.6161883044817767e-06, "loss": 0.6202, "step": 1950 }, { "epoch": 1.4850561583856843, "grad_norm": 32.20383071899414, "learning_rate": 1.6117050822515635e-06, "loss": 0.7464, "step": 1951 }, { "epoch": 1.4858176280220827, "grad_norm": 31.3388671875, "learning_rate": 1.6072268917251755e-06, "loss": 0.7226, "step": 1952 }, { "epoch": 1.4865790976584807, "grad_norm": 27.72353744506836, "learning_rate": 1.6027537395528403e-06, "loss": 0.7395, "step": 1953 }, { "epoch": 1.4873405672948792, "grad_norm": 26.9237060546875, "learning_rate": 1.5982856323773004e-06, "loss": 0.6078, "step": 1954 }, { "epoch": 1.4881020369312774, "grad_norm": 33.95270538330078, "learning_rate": 1.593822576833814e-06, "loss": 0.685, "step": 1955 }, { "epoch": 1.4888635065676756, "grad_norm": 25.47856330871582, "learning_rate": 1.5893645795501256e-06, "loss": 0.7515, "step": 1956 }, { "epoch": 1.4896249762040739, "grad_norm": 30.650344848632812, "learning_rate": 1.584911647146475e-06, "loss": 0.6852, "step": 1957 }, { "epoch": 1.490386445840472, "grad_norm": 24.075429916381836, "learning_rate": 1.5804637862355826e-06, "loss": 0.5456, "step": 1958 }, { "epoch": 1.4911479154768703, "grad_norm": 28.416305541992188, "learning_rate": 1.5760210034226352e-06, "loss": 0.6695, "step": 1959 }, { "epoch": 1.4919093851132685, "grad_norm": 28.5175724029541, "learning_rate": 1.5715833053052737e-06, "loss": 0.5108, "step": 1960 }, { "epoch": 1.492670854749667, "grad_norm": 32.036949157714844, "learning_rate": 1.567150698473593e-06, "loss": 0.7157, "step": 1961 }, { "epoch": 1.493432324386065, "grad_norm": 23.146739959716797, "learning_rate": 1.5627231895101307e-06, "loss": 0.7081, "step": 1962 }, { "epoch": 1.4941937940224634, "grad_norm": 31.808557510375977, "learning_rate": 1.5583007849898457e-06, "loss": 0.7053, "step": 1963 }, { "epoch": 1.4949552636588617, "grad_norm": 31.3686466217041, "learning_rate": 1.553883491480122e-06, "loss": 0.7005, "step": 1964 }, { "epoch": 1.49571673329526, "grad_norm": 29.857423782348633, "learning_rate": 1.5494713155407509e-06, "loss": 0.7203, "step": 1965 }, { "epoch": 1.4964782029316581, "grad_norm": 31.896163940429688, "learning_rate": 1.5450642637239306e-06, "loss": 0.6608, "step": 1966 }, { "epoch": 1.4972396725680563, "grad_norm": 32.89180374145508, "learning_rate": 1.5406623425742395e-06, "loss": 0.596, "step": 1967 }, { "epoch": 1.4980011422044546, "grad_norm": 27.081106185913086, "learning_rate": 1.5362655586286418e-06, "loss": 0.6395, "step": 1968 }, { "epoch": 1.4987626118408528, "grad_norm": 26.688373565673828, "learning_rate": 1.5318739184164771e-06, "loss": 0.6677, "step": 1969 }, { "epoch": 1.499524081477251, "grad_norm": 27.175487518310547, "learning_rate": 1.527487428459442e-06, "loss": 0.6703, "step": 1970 }, { "epoch": 1.5002855511136493, "grad_norm": 30.373645782470703, "learning_rate": 1.5231060952715825e-06, "loss": 0.6117, "step": 1971 }, { "epoch": 1.5010470207500477, "grad_norm": 25.80720329284668, "learning_rate": 1.5187299253592885e-06, "loss": 0.6508, "step": 1972 }, { "epoch": 1.5018084903864457, "grad_norm": 22.668949127197266, "learning_rate": 1.5143589252212876e-06, "loss": 0.5937, "step": 1973 }, { "epoch": 1.5025699600228442, "grad_norm": 22.085233688354492, "learning_rate": 1.5099931013486246e-06, "loss": 0.4673, "step": 1974 }, { "epoch": 1.5033314296592424, "grad_norm": 27.728803634643555, "learning_rate": 1.5056324602246558e-06, "loss": 0.7363, "step": 1975 }, { "epoch": 1.5040928992956406, "grad_norm": 26.116809844970703, "learning_rate": 1.5012770083250444e-06, "loss": 0.6801, "step": 1976 }, { "epoch": 1.5048543689320388, "grad_norm": 28.951690673828125, "learning_rate": 1.4969267521177506e-06, "loss": 0.715, "step": 1977 }, { "epoch": 1.505615838568437, "grad_norm": 31.423612594604492, "learning_rate": 1.4925816980630114e-06, "loss": 0.6941, "step": 1978 }, { "epoch": 1.5063773082048353, "grad_norm": 29.68526268005371, "learning_rate": 1.4882418526133414e-06, "loss": 0.7056, "step": 1979 }, { "epoch": 1.5071387778412335, "grad_norm": 29.2845458984375, "learning_rate": 1.4839072222135248e-06, "loss": 0.6909, "step": 1980 }, { "epoch": 1.507900247477632, "grad_norm": 31.847999572753906, "learning_rate": 1.4795778133005972e-06, "loss": 0.7154, "step": 1981 }, { "epoch": 1.50866171711403, "grad_norm": 30.829710006713867, "learning_rate": 1.475253632303838e-06, "loss": 0.7386, "step": 1982 }, { "epoch": 1.5094231867504284, "grad_norm": 33.253211975097656, "learning_rate": 1.4709346856447648e-06, "loss": 0.7438, "step": 1983 }, { "epoch": 1.5101846563868264, "grad_norm": 27.949495315551758, "learning_rate": 1.4666209797371262e-06, "loss": 0.6489, "step": 1984 }, { "epoch": 1.5109461260232249, "grad_norm": 27.150630950927734, "learning_rate": 1.462312520986885e-06, "loss": 0.6619, "step": 1985 }, { "epoch": 1.511707595659623, "grad_norm": 26.018659591674805, "learning_rate": 1.4580093157922088e-06, "loss": 0.4952, "step": 1986 }, { "epoch": 1.5124690652960213, "grad_norm": 33.99868392944336, "learning_rate": 1.4537113705434658e-06, "loss": 0.6838, "step": 1987 }, { "epoch": 1.5132305349324195, "grad_norm": 32.598087310791016, "learning_rate": 1.4494186916232168e-06, "loss": 0.7373, "step": 1988 }, { "epoch": 1.5139920045688178, "grad_norm": 33.54926681518555, "learning_rate": 1.445131285406199e-06, "loss": 0.7666, "step": 1989 }, { "epoch": 1.5147534742052162, "grad_norm": 31.553672790527344, "learning_rate": 1.440849158259316e-06, "loss": 0.6769, "step": 1990 }, { "epoch": 1.5155149438416142, "grad_norm": 39.39175033569336, "learning_rate": 1.4365723165416357e-06, "loss": 0.3931, "step": 1991 }, { "epoch": 1.5162764134780127, "grad_norm": 34.64937973022461, "learning_rate": 1.4323007666043786e-06, "loss": 0.7878, "step": 1992 }, { "epoch": 1.5170378831144107, "grad_norm": 34.25494384765625, "learning_rate": 1.4280345147909058e-06, "loss": 0.7708, "step": 1993 }, { "epoch": 1.5177993527508091, "grad_norm": 29.373470306396484, "learning_rate": 1.4237735674367037e-06, "loss": 0.6537, "step": 1994 }, { "epoch": 1.5185608223872074, "grad_norm": 29.2237548828125, "learning_rate": 1.4195179308693918e-06, "loss": 0.5641, "step": 1995 }, { "epoch": 1.5193222920236056, "grad_norm": 29.810876846313477, "learning_rate": 1.4152676114086989e-06, "loss": 0.699, "step": 1996 }, { "epoch": 1.5200837616600038, "grad_norm": 30.24208641052246, "learning_rate": 1.4110226153664537e-06, "loss": 0.6554, "step": 1997 }, { "epoch": 1.520845231296402, "grad_norm": 32.16893768310547, "learning_rate": 1.406782949046583e-06, "loss": 0.8156, "step": 1998 }, { "epoch": 1.5216067009328003, "grad_norm": 24.01704978942871, "learning_rate": 1.4025486187451015e-06, "loss": 0.546, "step": 1999 }, { "epoch": 1.5223681705691985, "grad_norm": 31.152587890625, "learning_rate": 1.3983196307500973e-06, "loss": 0.8026, "step": 2000 }, { "epoch": 1.523129640205597, "grad_norm": 27.331802368164062, "learning_rate": 1.3940959913417218e-06, "loss": 0.7041, "step": 2001 }, { "epoch": 1.523891109841995, "grad_norm": 31.586074829101562, "learning_rate": 1.3898777067921866e-06, "loss": 0.7171, "step": 2002 }, { "epoch": 1.5246525794783934, "grad_norm": 35.31203079223633, "learning_rate": 1.3856647833657532e-06, "loss": 0.8825, "step": 2003 }, { "epoch": 1.5254140491147914, "grad_norm": 32.24921417236328, "learning_rate": 1.381457227318721e-06, "loss": 0.7703, "step": 2004 }, { "epoch": 1.5261755187511898, "grad_norm": 28.814512252807617, "learning_rate": 1.377255044899412e-06, "loss": 0.6599, "step": 2005 }, { "epoch": 1.526936988387588, "grad_norm": 30.17646026611328, "learning_rate": 1.3730582423481786e-06, "loss": 0.7093, "step": 2006 }, { "epoch": 1.5276984580239863, "grad_norm": 29.831832885742188, "learning_rate": 1.3688668258973776e-06, "loss": 0.8312, "step": 2007 }, { "epoch": 1.5284599276603845, "grad_norm": 29.608247756958008, "learning_rate": 1.3646808017713698e-06, "loss": 0.7391, "step": 2008 }, { "epoch": 1.5292213972967827, "grad_norm": 24.51552963256836, "learning_rate": 1.3605001761865022e-06, "loss": 0.5446, "step": 2009 }, { "epoch": 1.5299828669331812, "grad_norm": 28.088403701782227, "learning_rate": 1.3563249553511149e-06, "loss": 0.574, "step": 2010 }, { "epoch": 1.5307443365695792, "grad_norm": 27.008441925048828, "learning_rate": 1.3521551454655164e-06, "loss": 0.7005, "step": 2011 }, { "epoch": 1.5315058062059776, "grad_norm": 34.253047943115234, "learning_rate": 1.3479907527219772e-06, "loss": 0.8393, "step": 2012 }, { "epoch": 1.5322672758423757, "grad_norm": 34.83329772949219, "learning_rate": 1.343831783304726e-06, "loss": 0.7919, "step": 2013 }, { "epoch": 1.533028745478774, "grad_norm": 29.464492797851562, "learning_rate": 1.3396782433899403e-06, "loss": 0.6937, "step": 2014 }, { "epoch": 1.5337902151151723, "grad_norm": 40.99031066894531, "learning_rate": 1.3355301391457332e-06, "loss": 0.7882, "step": 2015 }, { "epoch": 1.5345516847515706, "grad_norm": 25.138065338134766, "learning_rate": 1.3313874767321388e-06, "loss": 0.7383, "step": 2016 }, { "epoch": 1.5353131543879688, "grad_norm": 32.828277587890625, "learning_rate": 1.3272502623011219e-06, "loss": 0.6085, "step": 2017 }, { "epoch": 1.536074624024367, "grad_norm": 34.066165924072266, "learning_rate": 1.3231185019965482e-06, "loss": 0.7098, "step": 2018 }, { "epoch": 1.5368360936607652, "grad_norm": 24.82942008972168, "learning_rate": 1.3189922019541879e-06, "loss": 0.6559, "step": 2019 }, { "epoch": 1.5375975632971635, "grad_norm": 30.60334587097168, "learning_rate": 1.3148713683016966e-06, "loss": 0.5545, "step": 2020 }, { "epoch": 1.538359032933562, "grad_norm": 32.431236267089844, "learning_rate": 1.3107560071586224e-06, "loss": 0.9109, "step": 2021 }, { "epoch": 1.53912050256996, "grad_norm": 31.49688720703125, "learning_rate": 1.3066461246363777e-06, "loss": 0.8368, "step": 2022 }, { "epoch": 1.5398819722063584, "grad_norm": 29.243955612182617, "learning_rate": 1.3025417268382452e-06, "loss": 0.729, "step": 2023 }, { "epoch": 1.5406434418427564, "grad_norm": 29.06166648864746, "learning_rate": 1.298442819859354e-06, "loss": 0.7349, "step": 2024 }, { "epoch": 1.5414049114791548, "grad_norm": 30.27044105529785, "learning_rate": 1.29434940978669e-06, "loss": 0.7032, "step": 2025 }, { "epoch": 1.542166381115553, "grad_norm": 27.429973602294922, "learning_rate": 1.2902615026990688e-06, "loss": 0.6211, "step": 2026 }, { "epoch": 1.5429278507519513, "grad_norm": 34.60334014892578, "learning_rate": 1.2861791046671368e-06, "loss": 0.6797, "step": 2027 }, { "epoch": 1.5436893203883495, "grad_norm": 21.984588623046875, "learning_rate": 1.2821022217533586e-06, "loss": 0.5308, "step": 2028 }, { "epoch": 1.5444507900247477, "grad_norm": 29.423526763916016, "learning_rate": 1.2780308600120078e-06, "loss": 0.7333, "step": 2029 }, { "epoch": 1.5452122596611462, "grad_norm": 26.124752044677734, "learning_rate": 1.2739650254891616e-06, "loss": 0.6378, "step": 2030 }, { "epoch": 1.5459737292975442, "grad_norm": 29.322174072265625, "learning_rate": 1.2699047242226832e-06, "loss": 0.5448, "step": 2031 }, { "epoch": 1.5467351989339426, "grad_norm": 36.079166412353516, "learning_rate": 1.265849962242227e-06, "loss": 0.8169, "step": 2032 }, { "epoch": 1.5474966685703406, "grad_norm": 30.17820930480957, "learning_rate": 1.261800745569216e-06, "loss": 0.6212, "step": 2033 }, { "epoch": 1.548258138206739, "grad_norm": 35.503631591796875, "learning_rate": 1.2577570802168415e-06, "loss": 0.9922, "step": 2034 }, { "epoch": 1.5490196078431373, "grad_norm": 26.30396842956543, "learning_rate": 1.2537189721900434e-06, "loss": 0.575, "step": 2035 }, { "epoch": 1.5497810774795355, "grad_norm": 35.36972427368164, "learning_rate": 1.2496864274855198e-06, "loss": 0.8277, "step": 2036 }, { "epoch": 1.5505425471159338, "grad_norm": 31.002498626708984, "learning_rate": 1.2456594520916999e-06, "loss": 0.7411, "step": 2037 }, { "epoch": 1.551304016752332, "grad_norm": 27.736507415771484, "learning_rate": 1.2416380519887445e-06, "loss": 0.5911, "step": 2038 }, { "epoch": 1.5520654863887302, "grad_norm": 30.667980194091797, "learning_rate": 1.237622233148535e-06, "loss": 0.6101, "step": 2039 }, { "epoch": 1.5528269560251284, "grad_norm": 33.007057189941406, "learning_rate": 1.233612001534663e-06, "loss": 0.7593, "step": 2040 }, { "epoch": 1.5535884256615269, "grad_norm": 29.56964111328125, "learning_rate": 1.2296073631024253e-06, "loss": 0.7431, "step": 2041 }, { "epoch": 1.5543498952979249, "grad_norm": 26.447463989257812, "learning_rate": 1.225608323798811e-06, "loss": 0.7032, "step": 2042 }, { "epoch": 1.5551113649343233, "grad_norm": 30.827373504638672, "learning_rate": 1.2216148895624947e-06, "loss": 0.6749, "step": 2043 }, { "epoch": 1.5558728345707213, "grad_norm": 34.54490661621094, "learning_rate": 1.2176270663238276e-06, "loss": 0.6571, "step": 2044 }, { "epoch": 1.5566343042071198, "grad_norm": 25.419261932373047, "learning_rate": 1.21364486000483e-06, "loss": 0.6191, "step": 2045 }, { "epoch": 1.557395773843518, "grad_norm": 33.08875274658203, "learning_rate": 1.2096682765191742e-06, "loss": 0.729, "step": 2046 }, { "epoch": 1.5581572434799162, "grad_norm": 33.58367156982422, "learning_rate": 1.2056973217721924e-06, "loss": 0.6012, "step": 2047 }, { "epoch": 1.5589187131163145, "grad_norm": 25.962663650512695, "learning_rate": 1.2017320016608509e-06, "loss": 0.5267, "step": 2048 }, { "epoch": 1.5596801827527127, "grad_norm": 34.00405502319336, "learning_rate": 1.1977723220737513e-06, "loss": 0.7319, "step": 2049 }, { "epoch": 1.5604416523891111, "grad_norm": 27.631608963012695, "learning_rate": 1.1938182888911176e-06, "loss": 0.6683, "step": 2050 }, { "epoch": 1.5612031220255091, "grad_norm": 31.08919906616211, "learning_rate": 1.1898699079847898e-06, "loss": 0.7543, "step": 2051 }, { "epoch": 1.5619645916619076, "grad_norm": 26.516399383544922, "learning_rate": 1.1859271852182137e-06, "loss": 0.6638, "step": 2052 }, { "epoch": 1.5627260612983056, "grad_norm": 28.806747436523438, "learning_rate": 1.181990126446431e-06, "loss": 0.6314, "step": 2053 }, { "epoch": 1.563487530934704, "grad_norm": 33.17510223388672, "learning_rate": 1.1780587375160763e-06, "loss": 0.7168, "step": 2054 }, { "epoch": 1.5642490005711023, "grad_norm": 25.93999671936035, "learning_rate": 1.1741330242653598e-06, "loss": 0.6455, "step": 2055 }, { "epoch": 1.5650104702075005, "grad_norm": 32.30558776855469, "learning_rate": 1.1702129925240658e-06, "loss": 0.7601, "step": 2056 }, { "epoch": 1.5657719398438987, "grad_norm": 43.337860107421875, "learning_rate": 1.1662986481135407e-06, "loss": 0.8362, "step": 2057 }, { "epoch": 1.566533409480297, "grad_norm": 33.84115982055664, "learning_rate": 1.1623899968466857e-06, "loss": 0.6689, "step": 2058 }, { "epoch": 1.5672948791166952, "grad_norm": 35.26539993286133, "learning_rate": 1.1584870445279461e-06, "loss": 0.8394, "step": 2059 }, { "epoch": 1.5680563487530934, "grad_norm": 31.748699188232422, "learning_rate": 1.154589796953306e-06, "loss": 0.6578, "step": 2060 }, { "epoch": 1.5688178183894919, "grad_norm": 29.427682876586914, "learning_rate": 1.1506982599102756e-06, "loss": 0.6335, "step": 2061 }, { "epoch": 1.5695792880258899, "grad_norm": 32.036075592041016, "learning_rate": 1.1468124391778874e-06, "loss": 0.6998, "step": 2062 }, { "epoch": 1.5703407576622883, "grad_norm": 29.439409255981445, "learning_rate": 1.1429323405266824e-06, "loss": 0.8264, "step": 2063 }, { "epoch": 1.5711022272986863, "grad_norm": 31.589263916015625, "learning_rate": 1.1390579697187065e-06, "loss": 0.6886, "step": 2064 }, { "epoch": 1.5718636969350848, "grad_norm": 31.235139846801758, "learning_rate": 1.1351893325074975e-06, "loss": 0.6628, "step": 2065 }, { "epoch": 1.572625166571483, "grad_norm": 36.89226150512695, "learning_rate": 1.131326434638081e-06, "loss": 0.7956, "step": 2066 }, { "epoch": 1.5733866362078812, "grad_norm": 29.986570358276367, "learning_rate": 1.1274692818469568e-06, "loss": 0.745, "step": 2067 }, { "epoch": 1.5741481058442794, "grad_norm": 27.775447845458984, "learning_rate": 1.1236178798620956e-06, "loss": 0.6265, "step": 2068 }, { "epoch": 1.5749095754806777, "grad_norm": 30.107019424438477, "learning_rate": 1.1197722344029265e-06, "loss": 0.753, "step": 2069 }, { "epoch": 1.5756710451170761, "grad_norm": 27.705503463745117, "learning_rate": 1.1159323511803311e-06, "loss": 0.5101, "step": 2070 }, { "epoch": 1.5764325147534741, "grad_norm": 28.064908981323242, "learning_rate": 1.1120982358966326e-06, "loss": 0.6031, "step": 2071 }, { "epoch": 1.5771939843898726, "grad_norm": 36.96857833862305, "learning_rate": 1.1082698942455895e-06, "loss": 0.8703, "step": 2072 }, { "epoch": 1.5779554540262706, "grad_norm": 25.79486656188965, "learning_rate": 1.104447331912386e-06, "loss": 0.4833, "step": 2073 }, { "epoch": 1.578716923662669, "grad_norm": 28.773761749267578, "learning_rate": 1.100630554573624e-06, "loss": 0.6715, "step": 2074 }, { "epoch": 1.5794783932990673, "grad_norm": 31.56308937072754, "learning_rate": 1.0968195678973153e-06, "loss": 0.5641, "step": 2075 }, { "epoch": 1.5802398629354655, "grad_norm": 36.77650451660156, "learning_rate": 1.0930143775428714e-06, "loss": 0.6928, "step": 2076 }, { "epoch": 1.5810013325718637, "grad_norm": 29.2596435546875, "learning_rate": 1.0892149891610965e-06, "loss": 0.6128, "step": 2077 }, { "epoch": 1.581762802208262, "grad_norm": 38.17643356323242, "learning_rate": 1.0854214083941778e-06, "loss": 0.7472, "step": 2078 }, { "epoch": 1.5825242718446602, "grad_norm": 29.41641616821289, "learning_rate": 1.08163364087568e-06, "loss": 0.7222, "step": 2079 }, { "epoch": 1.5832857414810584, "grad_norm": 27.08153533935547, "learning_rate": 1.0778516922305331e-06, "loss": 0.5967, "step": 2080 }, { "epoch": 1.5840472111174568, "grad_norm": 30.507238388061523, "learning_rate": 1.074075568075027e-06, "loss": 0.7236, "step": 2081 }, { "epoch": 1.5848086807538548, "grad_norm": 29.182662963867188, "learning_rate": 1.0703052740168018e-06, "loss": 0.6062, "step": 2082 }, { "epoch": 1.5855701503902533, "grad_norm": 29.97278594970703, "learning_rate": 1.0665408156548395e-06, "loss": 0.6465, "step": 2083 }, { "epoch": 1.5863316200266513, "grad_norm": 36.84199905395508, "learning_rate": 1.0627821985794561e-06, "loss": 0.8333, "step": 2084 }, { "epoch": 1.5870930896630497, "grad_norm": 31.571657180786133, "learning_rate": 1.059029428372294e-06, "loss": 0.6432, "step": 2085 }, { "epoch": 1.587854559299448, "grad_norm": 27.03429412841797, "learning_rate": 1.0552825106063114e-06, "loss": 0.6748, "step": 2086 }, { "epoch": 1.5886160289358462, "grad_norm": 28.05469512939453, "learning_rate": 1.051541450845776e-06, "loss": 0.5929, "step": 2087 }, { "epoch": 1.5893774985722444, "grad_norm": 28.494518280029297, "learning_rate": 1.0478062546462569e-06, "loss": 0.6148, "step": 2088 }, { "epoch": 1.5901389682086426, "grad_norm": 26.496580123901367, "learning_rate": 1.0440769275546153e-06, "loss": 0.6345, "step": 2089 }, { "epoch": 1.590900437845041, "grad_norm": 29.254432678222656, "learning_rate": 1.0403534751089955e-06, "loss": 0.7671, "step": 2090 }, { "epoch": 1.591661907481439, "grad_norm": 29.0679874420166, "learning_rate": 1.036635902838819e-06, "loss": 0.6629, "step": 2091 }, { "epoch": 1.5924233771178375, "grad_norm": 29.54790496826172, "learning_rate": 1.0329242162647751e-06, "loss": 0.5981, "step": 2092 }, { "epoch": 1.5931848467542356, "grad_norm": 26.639490127563477, "learning_rate": 1.0292184208988133e-06, "loss": 0.5431, "step": 2093 }, { "epoch": 1.593946316390634, "grad_norm": 29.553543090820312, "learning_rate": 1.0255185222441317e-06, "loss": 0.548, "step": 2094 }, { "epoch": 1.5947077860270322, "grad_norm": 26.58224868774414, "learning_rate": 1.0218245257951753e-06, "loss": 0.605, "step": 2095 }, { "epoch": 1.5954692556634305, "grad_norm": 26.12196159362793, "learning_rate": 1.0181364370376212e-06, "loss": 0.6856, "step": 2096 }, { "epoch": 1.5962307252998287, "grad_norm": 28.62148666381836, "learning_rate": 1.014454261448376e-06, "loss": 0.593, "step": 2097 }, { "epoch": 1.596992194936227, "grad_norm": 25.782094955444336, "learning_rate": 1.0107780044955623e-06, "loss": 0.5485, "step": 2098 }, { "epoch": 1.5977536645726251, "grad_norm": 44.426334381103516, "learning_rate": 1.007107671638516e-06, "loss": 0.7846, "step": 2099 }, { "epoch": 1.5985151342090234, "grad_norm": 30.733142852783203, "learning_rate": 1.003443268327774e-06, "loss": 0.728, "step": 2100 }, { "epoch": 1.5992766038454218, "grad_norm": 24.533777236938477, "learning_rate": 9.99784800005068e-07, "loss": 0.435, "step": 2101 }, { "epoch": 1.6000380734818198, "grad_norm": 37.96542739868164, "learning_rate": 9.961322721033163e-07, "loss": 0.7428, "step": 2102 }, { "epoch": 1.6007995431182183, "grad_norm": 35.20521545410156, "learning_rate": 9.924856900466152e-07, "loss": 0.7831, "step": 2103 }, { "epoch": 1.6015610127546163, "grad_norm": 35.16457748413086, "learning_rate": 9.888450592502318e-07, "loss": 0.6417, "step": 2104 }, { "epoch": 1.6023224823910147, "grad_norm": 28.596393585205078, "learning_rate": 9.852103851205946e-07, "loss": 0.6217, "step": 2105 }, { "epoch": 1.603083952027413, "grad_norm": 34.43901062011719, "learning_rate": 9.81581673055288e-07, "loss": 0.9036, "step": 2106 }, { "epoch": 1.6038454216638112, "grad_norm": 26.20008659362793, "learning_rate": 9.779589284430396e-07, "loss": 0.5474, "step": 2107 }, { "epoch": 1.6046068913002094, "grad_norm": 23.07138442993164, "learning_rate": 9.743421566637168e-07, "loss": 0.5041, "step": 2108 }, { "epoch": 1.6053683609366076, "grad_norm": 30.84495735168457, "learning_rate": 9.707313630883214e-07, "loss": 0.7455, "step": 2109 }, { "epoch": 1.606129830573006, "grad_norm": 32.7132453918457, "learning_rate": 9.671265530789692e-07, "loss": 0.657, "step": 2110 }, { "epoch": 1.606891300209404, "grad_norm": 36.743446350097656, "learning_rate": 9.635277319888952e-07, "loss": 0.7413, "step": 2111 }, { "epoch": 1.6076527698458025, "grad_norm": 28.286968231201172, "learning_rate": 9.599349051624402e-07, "loss": 0.6724, "step": 2112 }, { "epoch": 1.6084142394822005, "grad_norm": 33.76932907104492, "learning_rate": 9.563480779350438e-07, "loss": 0.6656, "step": 2113 }, { "epoch": 1.609175709118599, "grad_norm": 44.41170120239258, "learning_rate": 9.527672556332345e-07, "loss": 0.8593, "step": 2114 }, { "epoch": 1.609937178754997, "grad_norm": 29.339557647705078, "learning_rate": 9.491924435746247e-07, "loss": 0.7353, "step": 2115 }, { "epoch": 1.6106986483913954, "grad_norm": 35.96495819091797, "learning_rate": 9.456236470679003e-07, "loss": 0.7716, "step": 2116 }, { "epoch": 1.6114601180277937, "grad_norm": 34.59695053100586, "learning_rate": 9.420608714128155e-07, "loss": 0.8038, "step": 2117 }, { "epoch": 1.6122215876641919, "grad_norm": 34.225521087646484, "learning_rate": 9.385041219001817e-07, "loss": 0.8303, "step": 2118 }, { "epoch": 1.61298305730059, "grad_norm": 34.64314270019531, "learning_rate": 9.349534038118608e-07, "loss": 0.7925, "step": 2119 }, { "epoch": 1.6137445269369883, "grad_norm": 30.03654670715332, "learning_rate": 9.31408722420763e-07, "loss": 0.6495, "step": 2120 }, { "epoch": 1.6145059965733868, "grad_norm": 27.334407806396484, "learning_rate": 9.278700829908266e-07, "loss": 0.5838, "step": 2121 }, { "epoch": 1.6152674662097848, "grad_norm": 33.08980178833008, "learning_rate": 9.24337490777022e-07, "loss": 0.8399, "step": 2122 }, { "epoch": 1.6160289358461832, "grad_norm": 29.446699142456055, "learning_rate": 9.208109510253366e-07, "loss": 0.6075, "step": 2123 }, { "epoch": 1.6167904054825812, "grad_norm": 42.0958366394043, "learning_rate": 9.172904689727752e-07, "loss": 0.8524, "step": 2124 }, { "epoch": 1.6175518751189797, "grad_norm": 35.52198791503906, "learning_rate": 9.137760498473391e-07, "loss": 0.8746, "step": 2125 }, { "epoch": 1.618313344755378, "grad_norm": 26.803062438964844, "learning_rate": 9.102676988680309e-07, "loss": 0.7315, "step": 2126 }, { "epoch": 1.6190748143917761, "grad_norm": 31.624183654785156, "learning_rate": 9.067654212448396e-07, "loss": 0.7031, "step": 2127 }, { "epoch": 1.6198362840281744, "grad_norm": 36.04194259643555, "learning_rate": 9.032692221787403e-07, "loss": 0.7392, "step": 2128 }, { "epoch": 1.6205977536645726, "grad_norm": 30.646621704101562, "learning_rate": 8.997791068616734e-07, "loss": 0.6374, "step": 2129 }, { "epoch": 1.6213592233009708, "grad_norm": 30.583839416503906, "learning_rate": 8.96295080476548e-07, "loss": 0.8087, "step": 2130 }, { "epoch": 1.622120692937369, "grad_norm": 37.712032318115234, "learning_rate": 8.928171481972359e-07, "loss": 0.8857, "step": 2131 }, { "epoch": 1.6228821625737675, "grad_norm": 35.08882141113281, "learning_rate": 8.893453151885507e-07, "loss": 0.866, "step": 2132 }, { "epoch": 1.6236436322101655, "grad_norm": 26.016958236694336, "learning_rate": 8.858795866062536e-07, "loss": 0.6957, "step": 2133 }, { "epoch": 1.624405101846564, "grad_norm": 29.985000610351562, "learning_rate": 8.824199675970385e-07, "loss": 0.584, "step": 2134 }, { "epoch": 1.625166571482962, "grad_norm": 30.87150001525879, "learning_rate": 8.789664632985306e-07, "loss": 0.6511, "step": 2135 }, { "epoch": 1.6259280411193604, "grad_norm": 38.97594451904297, "learning_rate": 8.755190788392681e-07, "loss": 0.8247, "step": 2136 }, { "epoch": 1.6266895107557586, "grad_norm": 32.37511444091797, "learning_rate": 8.720778193387047e-07, "loss": 0.7197, "step": 2137 }, { "epoch": 1.6274509803921569, "grad_norm": 24.394168853759766, "learning_rate": 8.686426899071976e-07, "loss": 0.6185, "step": 2138 }, { "epoch": 1.628212450028555, "grad_norm": 34.733158111572266, "learning_rate": 8.652136956460044e-07, "loss": 0.675, "step": 2139 }, { "epoch": 1.6289739196649533, "grad_norm": 30.3035831451416, "learning_rate": 8.617908416472647e-07, "loss": 0.6533, "step": 2140 }, { "epoch": 1.6297353893013518, "grad_norm": 28.481054306030273, "learning_rate": 8.583741329940049e-07, "loss": 0.646, "step": 2141 }, { "epoch": 1.6304968589377498, "grad_norm": 30.609439849853516, "learning_rate": 8.549635747601226e-07, "loss": 0.6988, "step": 2142 }, { "epoch": 1.6312583285741482, "grad_norm": 25.943239212036133, "learning_rate": 8.515591720103866e-07, "loss": 0.6364, "step": 2143 }, { "epoch": 1.6320197982105462, "grad_norm": 31.503292083740234, "learning_rate": 8.481609298004179e-07, "loss": 0.6195, "step": 2144 }, { "epoch": 1.6327812678469447, "grad_norm": 36.12900924682617, "learning_rate": 8.447688531766924e-07, "loss": 0.7339, "step": 2145 }, { "epoch": 1.6335427374833429, "grad_norm": 29.98458480834961, "learning_rate": 8.413829471765334e-07, "loss": 0.6243, "step": 2146 }, { "epoch": 1.6343042071197411, "grad_norm": 26.946714401245117, "learning_rate": 8.380032168280938e-07, "loss": 0.7377, "step": 2147 }, { "epoch": 1.6350656767561393, "grad_norm": 32.249366760253906, "learning_rate": 8.34629667150359e-07, "loss": 0.6776, "step": 2148 }, { "epoch": 1.6358271463925376, "grad_norm": 30.67563247680664, "learning_rate": 8.312623031531359e-07, "loss": 0.7443, "step": 2149 }, { "epoch": 1.6365886160289358, "grad_norm": 34.0343132019043, "learning_rate": 8.279011298370481e-07, "loss": 0.7956, "step": 2150 }, { "epoch": 1.637350085665334, "grad_norm": 35.368377685546875, "learning_rate": 8.2454615219352e-07, "loss": 0.7818, "step": 2151 }, { "epoch": 1.6381115553017325, "grad_norm": 30.426546096801758, "learning_rate": 8.211973752047792e-07, "loss": 0.6759, "step": 2152 }, { "epoch": 1.6388730249381305, "grad_norm": 28.504962921142578, "learning_rate": 8.178548038438433e-07, "loss": 0.6786, "step": 2153 }, { "epoch": 1.639634494574529, "grad_norm": 26.326515197753906, "learning_rate": 8.145184430745195e-07, "loss": 0.6874, "step": 2154 }, { "epoch": 1.640395964210927, "grad_norm": 28.690677642822266, "learning_rate": 8.111882978513841e-07, "loss": 0.734, "step": 2155 }, { "epoch": 1.6411574338473254, "grad_norm": 28.09772491455078, "learning_rate": 8.078643731197872e-07, "loss": 0.7516, "step": 2156 }, { "epoch": 1.6419189034837236, "grad_norm": 30.340097427368164, "learning_rate": 8.045466738158441e-07, "loss": 0.8082, "step": 2157 }, { "epoch": 1.6426803731201218, "grad_norm": 30.463314056396484, "learning_rate": 8.012352048664218e-07, "loss": 0.6018, "step": 2158 }, { "epoch": 1.64344184275652, "grad_norm": 35.2636833190918, "learning_rate": 7.97929971189133e-07, "loss": 0.8587, "step": 2159 }, { "epoch": 1.6442033123929183, "grad_norm": 29.126693725585938, "learning_rate": 7.946309776923345e-07, "loss": 0.6994, "step": 2160 }, { "epoch": 1.6449647820293167, "grad_norm": 32.419437408447266, "learning_rate": 7.913382292751171e-07, "loss": 0.6832, "step": 2161 }, { "epoch": 1.6457262516657147, "grad_norm": 29.627376556396484, "learning_rate": 7.880517308272955e-07, "loss": 0.6724, "step": 2162 }, { "epoch": 1.6464877213021132, "grad_norm": 37.55154800415039, "learning_rate": 7.847714872294005e-07, "loss": 0.7281, "step": 2163 }, { "epoch": 1.6472491909385112, "grad_norm": 31.11333465576172, "learning_rate": 7.814975033526778e-07, "loss": 0.8922, "step": 2164 }, { "epoch": 1.6480106605749096, "grad_norm": 26.14452362060547, "learning_rate": 7.782297840590797e-07, "loss": 0.6343, "step": 2165 }, { "epoch": 1.6487721302113079, "grad_norm": 43.487335205078125, "learning_rate": 7.749683342012481e-07, "loss": 0.8844, "step": 2166 }, { "epoch": 1.649533599847706, "grad_norm": 30.094945907592773, "learning_rate": 7.717131586225191e-07, "loss": 0.7403, "step": 2167 }, { "epoch": 1.6502950694841043, "grad_norm": 28.318283081054688, "learning_rate": 7.684642621569133e-07, "loss": 0.7516, "step": 2168 }, { "epoch": 1.6510565391205025, "grad_norm": 39.0335578918457, "learning_rate": 7.652216496291243e-07, "loss": 0.8246, "step": 2169 }, { "epoch": 1.6518180087569008, "grad_norm": 38.585472106933594, "learning_rate": 7.619853258545118e-07, "loss": 0.7692, "step": 2170 }, { "epoch": 1.652579478393299, "grad_norm": 27.209836959838867, "learning_rate": 7.587552956390981e-07, "loss": 0.6211, "step": 2171 }, { "epoch": 1.6533409480296974, "grad_norm": 37.91387176513672, "learning_rate": 7.555315637795624e-07, "loss": 0.7256, "step": 2172 }, { "epoch": 1.6541024176660954, "grad_norm": 40.89154052734375, "learning_rate": 7.523141350632285e-07, "loss": 0.9435, "step": 2173 }, { "epoch": 1.654863887302494, "grad_norm": 34.35085678100586, "learning_rate": 7.491030142680572e-07, "loss": 0.8343, "step": 2174 }, { "epoch": 1.655625356938892, "grad_norm": 26.987337112426758, "learning_rate": 7.458982061626441e-07, "loss": 0.6744, "step": 2175 }, { "epoch": 1.6563868265752903, "grad_norm": 37.9036865234375, "learning_rate": 7.426997155062126e-07, "loss": 0.8639, "step": 2176 }, { "epoch": 1.6571482962116886, "grad_norm": 28.941993713378906, "learning_rate": 7.39507547048603e-07, "loss": 0.6447, "step": 2177 }, { "epoch": 1.6579097658480868, "grad_norm": 39.62537384033203, "learning_rate": 7.363217055302624e-07, "loss": 0.888, "step": 2178 }, { "epoch": 1.658671235484485, "grad_norm": 30.14093589782715, "learning_rate": 7.331421956822493e-07, "loss": 0.7074, "step": 2179 }, { "epoch": 1.6594327051208833, "grad_norm": 32.665489196777344, "learning_rate": 7.299690222262174e-07, "loss": 0.6316, "step": 2180 }, { "epoch": 1.6601941747572817, "grad_norm": 28.16049575805664, "learning_rate": 7.268021898744066e-07, "loss": 0.674, "step": 2181 }, { "epoch": 1.6609556443936797, "grad_norm": 46.486629486083984, "learning_rate": 7.236417033296433e-07, "loss": 0.5978, "step": 2182 }, { "epoch": 1.6617171140300782, "grad_norm": 32.930328369140625, "learning_rate": 7.204875672853335e-07, "loss": 0.6529, "step": 2183 }, { "epoch": 1.6624785836664762, "grad_norm": 30.730792999267578, "learning_rate": 7.173397864254472e-07, "loss": 0.767, "step": 2184 }, { "epoch": 1.6632400533028746, "grad_norm": 34.0584602355957, "learning_rate": 7.141983654245183e-07, "loss": 0.7721, "step": 2185 }, { "epoch": 1.6640015229392728, "grad_norm": 26.838876724243164, "learning_rate": 7.110633089476354e-07, "loss": 0.547, "step": 2186 }, { "epoch": 1.664762992575671, "grad_norm": 25.050161361694336, "learning_rate": 7.079346216504396e-07, "loss": 0.6289, "step": 2187 }, { "epoch": 1.6655244622120693, "grad_norm": 29.258411407470703, "learning_rate": 7.048123081791108e-07, "loss": 0.5721, "step": 2188 }, { "epoch": 1.6662859318484675, "grad_norm": 29.47585105895996, "learning_rate": 7.016963731703602e-07, "loss": 0.5204, "step": 2189 }, { "epoch": 1.6670474014848657, "grad_norm": 27.289371490478516, "learning_rate": 6.985868212514335e-07, "loss": 0.6718, "step": 2190 }, { "epoch": 1.667808871121264, "grad_norm": 29.775022506713867, "learning_rate": 6.95483657040093e-07, "loss": 0.6948, "step": 2191 }, { "epoch": 1.6685703407576624, "grad_norm": 33.380638122558594, "learning_rate": 6.923868851446181e-07, "loss": 0.7087, "step": 2192 }, { "epoch": 1.6693318103940604, "grad_norm": 27.83628273010254, "learning_rate": 6.892965101637888e-07, "loss": 0.6966, "step": 2193 }, { "epoch": 1.6700932800304589, "grad_norm": 28.920995712280273, "learning_rate": 6.86212536686895e-07, "loss": 0.8073, "step": 2194 }, { "epoch": 1.6708547496668569, "grad_norm": 23.20146942138672, "learning_rate": 6.831349692937134e-07, "loss": 0.5757, "step": 2195 }, { "epoch": 1.6716162193032553, "grad_norm": 32.29258346557617, "learning_rate": 6.800638125545112e-07, "loss": 0.6358, "step": 2196 }, { "epoch": 1.6723776889396536, "grad_norm": 29.498416900634766, "learning_rate": 6.769990710300306e-07, "loss": 0.575, "step": 2197 }, { "epoch": 1.6731391585760518, "grad_norm": 29.090120315551758, "learning_rate": 6.739407492714934e-07, "loss": 0.6603, "step": 2198 }, { "epoch": 1.67390062821245, "grad_norm": 34.33234786987305, "learning_rate": 6.708888518205847e-07, "loss": 0.7696, "step": 2199 }, { "epoch": 1.6746620978488482, "grad_norm": 30.480268478393555, "learning_rate": 6.678433832094472e-07, "loss": 0.9015, "step": 2200 }, { "epoch": 1.6754235674852467, "grad_norm": 28.987363815307617, "learning_rate": 6.64804347960682e-07, "loss": 0.5358, "step": 2201 }, { "epoch": 1.6761850371216447, "grad_norm": 30.797557830810547, "learning_rate": 6.617717505873322e-07, "loss": 0.6385, "step": 2202 }, { "epoch": 1.6769465067580431, "grad_norm": 36.12929916381836, "learning_rate": 6.587455955928824e-07, "loss": 0.8537, "step": 2203 }, { "epoch": 1.6777079763944411, "grad_norm": 28.420392990112305, "learning_rate": 6.557258874712469e-07, "loss": 0.725, "step": 2204 }, { "epoch": 1.6784694460308396, "grad_norm": 27.9139347076416, "learning_rate": 6.527126307067716e-07, "loss": 0.6454, "step": 2205 }, { "epoch": 1.6792309156672378, "grad_norm": 35.29081344604492, "learning_rate": 6.497058297742182e-07, "loss": 0.6721, "step": 2206 }, { "epoch": 1.679992385303636, "grad_norm": 32.69925308227539, "learning_rate": 6.46705489138763e-07, "loss": 0.7451, "step": 2207 }, { "epoch": 1.6807538549400343, "grad_norm": 29.545547485351562, "learning_rate": 6.437116132559856e-07, "loss": 0.7039, "step": 2208 }, { "epoch": 1.6815153245764325, "grad_norm": 28.12919044494629, "learning_rate": 6.407242065718705e-07, "loss": 0.5818, "step": 2209 }, { "epoch": 1.6822767942128307, "grad_norm": 28.38266944885254, "learning_rate": 6.377432735227906e-07, "loss": 0.5849, "step": 2210 }, { "epoch": 1.683038263849229, "grad_norm": 24.897022247314453, "learning_rate": 6.347688185355072e-07, "loss": 0.5137, "step": 2211 }, { "epoch": 1.6837997334856274, "grad_norm": 32.73811340332031, "learning_rate": 6.318008460271618e-07, "loss": 0.8381, "step": 2212 }, { "epoch": 1.6845612031220254, "grad_norm": 27.208660125732422, "learning_rate": 6.288393604052689e-07, "loss": 0.6757, "step": 2213 }, { "epoch": 1.6853226727584238, "grad_norm": 32.94672393798828, "learning_rate": 6.258843660677105e-07, "loss": 0.8025, "step": 2214 }, { "epoch": 1.6860841423948218, "grad_norm": 29.2930908203125, "learning_rate": 6.229358674027241e-07, "loss": 0.7566, "step": 2215 }, { "epoch": 1.6868456120312203, "grad_norm": 28.089357376098633, "learning_rate": 6.199938687889095e-07, "loss": 0.687, "step": 2216 }, { "epoch": 1.6876070816676185, "grad_norm": 25.929040908813477, "learning_rate": 6.170583745952075e-07, "loss": 0.5608, "step": 2217 }, { "epoch": 1.6883685513040168, "grad_norm": 24.413536071777344, "learning_rate": 6.14129389180902e-07, "loss": 0.5854, "step": 2218 }, { "epoch": 1.689130020940415, "grad_norm": 29.321487426757812, "learning_rate": 6.11206916895608e-07, "loss": 0.7414, "step": 2219 }, { "epoch": 1.6898914905768132, "grad_norm": 32.374656677246094, "learning_rate": 6.082909620792743e-07, "loss": 0.7013, "step": 2220 }, { "epoch": 1.6906529602132117, "grad_norm": 31.762496948242188, "learning_rate": 6.053815290621662e-07, "loss": 0.678, "step": 2221 }, { "epoch": 1.6914144298496097, "grad_norm": 26.752973556518555, "learning_rate": 6.024786221648665e-07, "loss": 0.7752, "step": 2222 }, { "epoch": 1.692175899486008, "grad_norm": 31.72564125061035, "learning_rate": 5.995822456982625e-07, "loss": 0.7535, "step": 2223 }, { "epoch": 1.692937369122406, "grad_norm": 32.8385124206543, "learning_rate": 5.966924039635496e-07, "loss": 0.8685, "step": 2224 }, { "epoch": 1.6936988387588046, "grad_norm": 26.042144775390625, "learning_rate": 5.938091012522151e-07, "loss": 0.7102, "step": 2225 }, { "epoch": 1.6944603083952028, "grad_norm": 33.57801055908203, "learning_rate": 5.909323418460367e-07, "loss": 0.8879, "step": 2226 }, { "epoch": 1.695221778031601, "grad_norm": 30.9910831451416, "learning_rate": 5.880621300170752e-07, "loss": 0.6578, "step": 2227 }, { "epoch": 1.6959832476679992, "grad_norm": 34.433616638183594, "learning_rate": 5.851984700276675e-07, "loss": 0.6697, "step": 2228 }, { "epoch": 1.6967447173043975, "grad_norm": 32.08195495605469, "learning_rate": 5.823413661304211e-07, "loss": 0.7804, "step": 2229 }, { "epoch": 1.6975061869407957, "grad_norm": 38.08897399902344, "learning_rate": 5.794908225682083e-07, "loss": 0.9119, "step": 2230 }, { "epoch": 1.698267656577194, "grad_norm": 32.446311950683594, "learning_rate": 5.766468435741584e-07, "loss": 0.8485, "step": 2231 }, { "epoch": 1.6990291262135924, "grad_norm": 32.76416015625, "learning_rate": 5.73809433371652e-07, "loss": 0.7385, "step": 2232 }, { "epoch": 1.6997905958499904, "grad_norm": 36.371910095214844, "learning_rate": 5.709785961743169e-07, "loss": 0.571, "step": 2233 }, { "epoch": 1.7005520654863888, "grad_norm": 30.387042999267578, "learning_rate": 5.681543361860137e-07, "loss": 0.7301, "step": 2234 }, { "epoch": 1.7013135351227868, "grad_norm": 33.310546875, "learning_rate": 5.653366576008445e-07, "loss": 0.6413, "step": 2235 }, { "epoch": 1.7020750047591853, "grad_norm": 31.31390380859375, "learning_rate": 5.625255646031314e-07, "loss": 0.7461, "step": 2236 }, { "epoch": 1.7028364743955835, "grad_norm": 30.112937927246094, "learning_rate": 5.597210613674193e-07, "loss": 0.7922, "step": 2237 }, { "epoch": 1.7035979440319817, "grad_norm": 48.957763671875, "learning_rate": 5.569231520584662e-07, "loss": 0.902, "step": 2238 }, { "epoch": 1.70435941366838, "grad_norm": 33.73064422607422, "learning_rate": 5.541318408312385e-07, "loss": 0.7476, "step": 2239 }, { "epoch": 1.7051208833047782, "grad_norm": 34.50392150878906, "learning_rate": 5.513471318309044e-07, "loss": 0.6691, "step": 2240 }, { "epoch": 1.7058823529411766, "grad_norm": 27.21174430847168, "learning_rate": 5.48569029192827e-07, "loss": 0.6281, "step": 2241 }, { "epoch": 1.7066438225775746, "grad_norm": 27.56557273864746, "learning_rate": 5.457975370425588e-07, "loss": 0.6572, "step": 2242 }, { "epoch": 1.707405292213973, "grad_norm": 26.56990623474121, "learning_rate": 5.430326594958368e-07, "loss": 0.4974, "step": 2243 }, { "epoch": 1.708166761850371, "grad_norm": 24.55322265625, "learning_rate": 5.402744006585737e-07, "loss": 0.6303, "step": 2244 }, { "epoch": 1.7089282314867695, "grad_norm": 33.991695404052734, "learning_rate": 5.375227646268539e-07, "loss": 0.7902, "step": 2245 }, { "epoch": 1.7096897011231678, "grad_norm": 31.030162811279297, "learning_rate": 5.34777755486926e-07, "loss": 0.7599, "step": 2246 }, { "epoch": 1.710451170759566, "grad_norm": 27.946945190429688, "learning_rate": 5.320393773151993e-07, "loss": 0.7159, "step": 2247 }, { "epoch": 1.7112126403959642, "grad_norm": 28.079124450683594, "learning_rate": 5.293076341782339e-07, "loss": 0.6743, "step": 2248 }, { "epoch": 1.7119741100323624, "grad_norm": 23.430072784423828, "learning_rate": 5.265825301327376e-07, "loss": 0.5351, "step": 2249 }, { "epoch": 1.7127355796687607, "grad_norm": 26.333179473876953, "learning_rate": 5.238640692255598e-07, "loss": 0.6584, "step": 2250 }, { "epoch": 1.713497049305159, "grad_norm": 28.39971923828125, "learning_rate": 5.211522554936832e-07, "loss": 0.625, "step": 2251 }, { "epoch": 1.7142585189415573, "grad_norm": 30.517608642578125, "learning_rate": 5.184470929642199e-07, "loss": 0.7745, "step": 2252 }, { "epoch": 1.7150199885779553, "grad_norm": 31.25023078918457, "learning_rate": 5.157485856544048e-07, "loss": 0.646, "step": 2253 }, { "epoch": 1.7157814582143538, "grad_norm": 33.52351760864258, "learning_rate": 5.1305673757159e-07, "loss": 0.7319, "step": 2254 }, { "epoch": 1.7165429278507518, "grad_norm": 30.649023056030273, "learning_rate": 5.10371552713238e-07, "loss": 0.6784, "step": 2255 }, { "epoch": 1.7173043974871502, "grad_norm": 28.712512969970703, "learning_rate": 5.076930350669162e-07, "loss": 0.6559, "step": 2256 }, { "epoch": 1.7180658671235485, "grad_norm": 26.7255859375, "learning_rate": 5.050211886102913e-07, "loss": 0.6872, "step": 2257 }, { "epoch": 1.7188273367599467, "grad_norm": 29.03633689880371, "learning_rate": 5.02356017311123e-07, "loss": 0.6758, "step": 2258 }, { "epoch": 1.719588806396345, "grad_norm": 34.795318603515625, "learning_rate": 4.996975251272579e-07, "loss": 0.5713, "step": 2259 }, { "epoch": 1.7203502760327432, "grad_norm": 29.60315704345703, "learning_rate": 4.97045716006625e-07, "loss": 0.7226, "step": 2260 }, { "epoch": 1.7211117456691416, "grad_norm": 31.56952476501465, "learning_rate": 4.94400593887227e-07, "loss": 0.6504, "step": 2261 }, { "epoch": 1.7218732153055396, "grad_norm": 40.32448959350586, "learning_rate": 4.917621626971386e-07, "loss": 0.941, "step": 2262 }, { "epoch": 1.722634684941938, "grad_norm": 25.322059631347656, "learning_rate": 4.891304263544955e-07, "loss": 0.5701, "step": 2263 }, { "epoch": 1.723396154578336, "grad_norm": 29.843997955322266, "learning_rate": 4.865053887674942e-07, "loss": 0.6433, "step": 2264 }, { "epoch": 1.7241576242147345, "grad_norm": 30.76515769958496, "learning_rate": 4.838870538343815e-07, "loss": 0.655, "step": 2265 }, { "epoch": 1.7249190938511327, "grad_norm": 33.039093017578125, "learning_rate": 4.812754254434509e-07, "loss": 0.6527, "step": 2266 }, { "epoch": 1.725680563487531, "grad_norm": 25.238677978515625, "learning_rate": 4.786705074730374e-07, "loss": 0.713, "step": 2267 }, { "epoch": 1.7264420331239292, "grad_norm": 29.24014663696289, "learning_rate": 4.7607230379150935e-07, "loss": 0.7272, "step": 2268 }, { "epoch": 1.7272035027603274, "grad_norm": 31.178369522094727, "learning_rate": 4.734808182572659e-07, "loss": 0.6365, "step": 2269 }, { "epoch": 1.7279649723967256, "grad_norm": 34.92178726196289, "learning_rate": 4.70896054718728e-07, "loss": 0.7129, "step": 2270 }, { "epoch": 1.7287264420331239, "grad_norm": 30.77577018737793, "learning_rate": 4.683180170143359e-07, "loss": 0.6539, "step": 2271 }, { "epoch": 1.7294879116695223, "grad_norm": 32.83926773071289, "learning_rate": 4.657467089725398e-07, "loss": 0.6248, "step": 2272 }, { "epoch": 1.7302493813059203, "grad_norm": 32.05132293701172, "learning_rate": 4.6318213441179815e-07, "loss": 0.8676, "step": 2273 }, { "epoch": 1.7310108509423188, "grad_norm": 39.29285430908203, "learning_rate": 4.606242971405689e-07, "loss": 0.7102, "step": 2274 }, { "epoch": 1.7317723205787168, "grad_norm": 31.379751205444336, "learning_rate": 4.5807320095730577e-07, "loss": 0.7359, "step": 2275 }, { "epoch": 1.7325337902151152, "grad_norm": 30.045379638671875, "learning_rate": 4.5552884965045006e-07, "loss": 0.7099, "step": 2276 }, { "epoch": 1.7332952598515134, "grad_norm": 39.86351776123047, "learning_rate": 4.5299124699842876e-07, "loss": 0.7469, "step": 2277 }, { "epoch": 1.7340567294879117, "grad_norm": 30.267927169799805, "learning_rate": 4.5046039676964636e-07, "loss": 0.5808, "step": 2278 }, { "epoch": 1.73481819912431, "grad_norm": 25.970672607421875, "learning_rate": 4.4793630272247866e-07, "loss": 0.7431, "step": 2279 }, { "epoch": 1.7355796687607081, "grad_norm": 26.493732452392578, "learning_rate": 4.454189686052701e-07, "loss": 0.5343, "step": 2280 }, { "epoch": 1.7363411383971066, "grad_norm": 26.40505599975586, "learning_rate": 4.429083981563254e-07, "loss": 0.6655, "step": 2281 }, { "epoch": 1.7371026080335046, "grad_norm": 37.678401947021484, "learning_rate": 4.404045951039054e-07, "loss": 0.8539, "step": 2282 }, { "epoch": 1.737864077669903, "grad_norm": 24.892911911010742, "learning_rate": 4.379075631662205e-07, "loss": 0.4393, "step": 2283 }, { "epoch": 1.738625547306301, "grad_norm": 23.996042251586914, "learning_rate": 4.3541730605142775e-07, "loss": 0.5425, "step": 2284 }, { "epoch": 1.7393870169426995, "grad_norm": 32.724334716796875, "learning_rate": 4.3293382745762057e-07, "loss": 0.8369, "step": 2285 }, { "epoch": 1.7401484865790977, "grad_norm": 32.173011779785156, "learning_rate": 4.304571310728284e-07, "loss": 0.8261, "step": 2286 }, { "epoch": 1.740909956215496, "grad_norm": 28.027666091918945, "learning_rate": 4.27987220575008e-07, "loss": 0.6259, "step": 2287 }, { "epoch": 1.7416714258518942, "grad_norm": 34.16132354736328, "learning_rate": 4.255240996320392e-07, "loss": 0.6467, "step": 2288 }, { "epoch": 1.7424328954882924, "grad_norm": 27.693716049194336, "learning_rate": 4.23067771901719e-07, "loss": 0.5927, "step": 2289 }, { "epoch": 1.7431943651246906, "grad_norm": 31.316354751586914, "learning_rate": 4.206182410317561e-07, "loss": 0.7006, "step": 2290 }, { "epoch": 1.7439558347610888, "grad_norm": 28.683517456054688, "learning_rate": 4.181755106597657e-07, "loss": 0.709, "step": 2291 }, { "epoch": 1.7447173043974873, "grad_norm": 32.50971221923828, "learning_rate": 4.15739584413265e-07, "loss": 0.5897, "step": 2292 }, { "epoch": 1.7454787740338853, "grad_norm": 20.653600692749023, "learning_rate": 4.1331046590966593e-07, "loss": 0.399, "step": 2293 }, { "epoch": 1.7462402436702837, "grad_norm": 27.78783416748047, "learning_rate": 4.1088815875627076e-07, "loss": 0.5524, "step": 2294 }, { "epoch": 1.7470017133066817, "grad_norm": 28.15583038330078, "learning_rate": 4.0847266655026753e-07, "loss": 0.6961, "step": 2295 }, { "epoch": 1.7477631829430802, "grad_norm": 26.51068115234375, "learning_rate": 4.060639928787219e-07, "loss": 0.6432, "step": 2296 }, { "epoch": 1.7485246525794784, "grad_norm": 33.15547180175781, "learning_rate": 4.036621413185782e-07, "loss": 0.7731, "step": 2297 }, { "epoch": 1.7492861222158766, "grad_norm": 30.943445205688477, "learning_rate": 4.0126711543664433e-07, "loss": 0.633, "step": 2298 }, { "epoch": 1.7500475918522749, "grad_norm": 32.601531982421875, "learning_rate": 3.988789187895953e-07, "loss": 0.6289, "step": 2299 }, { "epoch": 1.750809061488673, "grad_norm": 26.338903427124023, "learning_rate": 3.9649755492396414e-07, "loss": 0.8332, "step": 2300 }, { "epoch": 1.7515705311250716, "grad_norm": 32.43155288696289, "learning_rate": 3.941230273761354e-07, "loss": 0.7203, "step": 2301 }, { "epoch": 1.7523320007614696, "grad_norm": 30.228130340576172, "learning_rate": 3.9175533967234335e-07, "loss": 0.7029, "step": 2302 }, { "epoch": 1.753093470397868, "grad_norm": 30.615478515625, "learning_rate": 3.8939449532866437e-07, "loss": 0.743, "step": 2303 }, { "epoch": 1.753854940034266, "grad_norm": 37.94570541381836, "learning_rate": 3.8704049785101173e-07, "loss": 0.8863, "step": 2304 }, { "epoch": 1.7546164096706645, "grad_norm": 32.52972412109375, "learning_rate": 3.846933507351319e-07, "loss": 0.7181, "step": 2305 }, { "epoch": 1.7553778793070627, "grad_norm": 34.856689453125, "learning_rate": 3.8235305746659737e-07, "loss": 0.809, "step": 2306 }, { "epoch": 1.756139348943461, "grad_norm": 33.77072525024414, "learning_rate": 3.8001962152080185e-07, "loss": 0.6255, "step": 2307 }, { "epoch": 1.7569008185798591, "grad_norm": 34.31856155395508, "learning_rate": 3.776930463629602e-07, "loss": 0.6658, "step": 2308 }, { "epoch": 1.7576622882162574, "grad_norm": 28.501949310302734, "learning_rate": 3.7537333544809264e-07, "loss": 0.7534, "step": 2309 }, { "epoch": 1.7584237578526556, "grad_norm": 27.88667106628418, "learning_rate": 3.7306049222102903e-07, "loss": 0.6826, "step": 2310 }, { "epoch": 1.7591852274890538, "grad_norm": 34.051998138427734, "learning_rate": 3.707545201163998e-07, "loss": 0.801, "step": 2311 }, { "epoch": 1.7599466971254523, "grad_norm": 37.63811111450195, "learning_rate": 3.6845542255863456e-07, "loss": 0.9395, "step": 2312 }, { "epoch": 1.7607081667618503, "grad_norm": 25.479955673217773, "learning_rate": 3.6616320296194886e-07, "loss": 0.6044, "step": 2313 }, { "epoch": 1.7614696363982487, "grad_norm": 31.365942001342773, "learning_rate": 3.638778647303476e-07, "loss": 0.6635, "step": 2314 }, { "epoch": 1.7622311060346467, "grad_norm": 29.502397537231445, "learning_rate": 3.61599411257616e-07, "loss": 0.6535, "step": 2315 }, { "epoch": 1.7629925756710452, "grad_norm": 39.46318435668945, "learning_rate": 3.593278459273153e-07, "loss": 0.819, "step": 2316 }, { "epoch": 1.7637540453074434, "grad_norm": 33.56135940551758, "learning_rate": 3.5706317211277763e-07, "loss": 0.7893, "step": 2317 }, { "epoch": 1.7645155149438416, "grad_norm": 29.62611198425293, "learning_rate": 3.548053931770995e-07, "loss": 0.5972, "step": 2318 }, { "epoch": 1.7652769845802398, "grad_norm": 31.55545997619629, "learning_rate": 3.525545124731439e-07, "loss": 0.6726, "step": 2319 }, { "epoch": 1.766038454216638, "grad_norm": 30.696300506591797, "learning_rate": 3.503105333435214e-07, "loss": 0.5263, "step": 2320 }, { "epoch": 1.7667999238530365, "grad_norm": 36.31211853027344, "learning_rate": 3.480734591205998e-07, "loss": 0.8577, "step": 2321 }, { "epoch": 1.7675613934894345, "grad_norm": 39.10386657714844, "learning_rate": 3.458432931264899e-07, "loss": 0.7221, "step": 2322 }, { "epoch": 1.768322863125833, "grad_norm": 34.17140579223633, "learning_rate": 3.436200386730476e-07, "loss": 0.8568, "step": 2323 }, { "epoch": 1.769084332762231, "grad_norm": 28.56905746459961, "learning_rate": 3.4140369906185913e-07, "loss": 0.6762, "step": 2324 }, { "epoch": 1.7698458023986294, "grad_norm": 35.65485382080078, "learning_rate": 3.3919427758424607e-07, "loss": 0.8124, "step": 2325 }, { "epoch": 1.7706072720350277, "grad_norm": 33.530677795410156, "learning_rate": 3.369917775212539e-07, "loss": 0.7913, "step": 2326 }, { "epoch": 1.7713687416714259, "grad_norm": 30.015727996826172, "learning_rate": 3.347962021436546e-07, "loss": 0.7803, "step": 2327 }, { "epoch": 1.772130211307824, "grad_norm": 35.93435287475586, "learning_rate": 3.3260755471192953e-07, "loss": 0.7429, "step": 2328 }, { "epoch": 1.7728916809442223, "grad_norm": 29.3546142578125, "learning_rate": 3.3042583847627653e-07, "loss": 0.5722, "step": 2329 }, { "epoch": 1.7736531505806206, "grad_norm": 28.38958740234375, "learning_rate": 3.282510566766012e-07, "loss": 0.7395, "step": 2330 }, { "epoch": 1.7744146202170188, "grad_norm": 26.225753784179688, "learning_rate": 3.2608321254250896e-07, "loss": 0.6705, "step": 2331 }, { "epoch": 1.7751760898534172, "grad_norm": 33.50041961669922, "learning_rate": 3.239223092933019e-07, "loss": 0.6769, "step": 2332 }, { "epoch": 1.7759375594898152, "grad_norm": 29.13167953491211, "learning_rate": 3.217683501379765e-07, "loss": 0.4939, "step": 2333 }, { "epoch": 1.7766990291262137, "grad_norm": 30.411380767822266, "learning_rate": 3.1962133827521835e-07, "loss": 0.7396, "step": 2334 }, { "epoch": 1.7774604987626117, "grad_norm": 35.63988494873047, "learning_rate": 3.174812768933927e-07, "loss": 0.7405, "step": 2335 }, { "epoch": 1.7782219683990101, "grad_norm": 29.541711807250977, "learning_rate": 3.153481691705457e-07, "loss": 0.651, "step": 2336 }, { "epoch": 1.7789834380354084, "grad_norm": 37.87580871582031, "learning_rate": 3.132220182743956e-07, "loss": 0.7073, "step": 2337 }, { "epoch": 1.7797449076718066, "grad_norm": 34.444549560546875, "learning_rate": 3.1110282736233233e-07, "loss": 0.719, "step": 2338 }, { "epoch": 1.7805063773082048, "grad_norm": 32.19717788696289, "learning_rate": 3.0899059958140576e-07, "loss": 0.7846, "step": 2339 }, { "epoch": 1.781267846944603, "grad_norm": 37.63265609741211, "learning_rate": 3.0688533806832845e-07, "loss": 0.6457, "step": 2340 }, { "epoch": 1.7820293165810015, "grad_norm": 24.643903732299805, "learning_rate": 3.0478704594946786e-07, "loss": 0.5322, "step": 2341 }, { "epoch": 1.7827907862173995, "grad_norm": 28.00391387939453, "learning_rate": 3.0269572634084144e-07, "loss": 0.6553, "step": 2342 }, { "epoch": 1.783552255853798, "grad_norm": 31.84063148498535, "learning_rate": 3.006113823481099e-07, "loss": 0.6793, "step": 2343 }, { "epoch": 1.784313725490196, "grad_norm": 29.44065284729004, "learning_rate": 2.985340170665768e-07, "loss": 0.7105, "step": 2344 }, { "epoch": 1.7850751951265944, "grad_norm": 28.025999069213867, "learning_rate": 2.9646363358118434e-07, "loss": 0.728, "step": 2345 }, { "epoch": 1.7858366647629924, "grad_norm": 32.748863220214844, "learning_rate": 2.944002349665043e-07, "loss": 0.7113, "step": 2346 }, { "epoch": 1.7865981343993909, "grad_norm": 33.03827667236328, "learning_rate": 2.9234382428673435e-07, "loss": 0.6073, "step": 2347 }, { "epoch": 1.787359604035789, "grad_norm": 48.89390563964844, "learning_rate": 2.902944045956962e-07, "loss": 0.8496, "step": 2348 }, { "epoch": 1.7881210736721873, "grad_norm": 33.290367126464844, "learning_rate": 2.8825197893683155e-07, "loss": 0.7904, "step": 2349 }, { "epoch": 1.7888825433085855, "grad_norm": 25.7712345123291, "learning_rate": 2.862165503431946e-07, "loss": 0.5494, "step": 2350 }, { "epoch": 1.7896440129449838, "grad_norm": 34.79444885253906, "learning_rate": 2.8418812183744593e-07, "loss": 0.6829, "step": 2351 }, { "epoch": 1.7904054825813822, "grad_norm": 40.69157028198242, "learning_rate": 2.821666964318553e-07, "loss": 0.8467, "step": 2352 }, { "epoch": 1.7911669522177802, "grad_norm": 31.949752807617188, "learning_rate": 2.8015227712829106e-07, "loss": 0.7077, "step": 2353 }, { "epoch": 1.7919284218541787, "grad_norm": 21.888437271118164, "learning_rate": 2.7814486691821574e-07, "loss": 0.5851, "step": 2354 }, { "epoch": 1.7926898914905767, "grad_norm": 25.899770736694336, "learning_rate": 2.7614446878268363e-07, "loss": 0.6959, "step": 2355 }, { "epoch": 1.7934513611269751, "grad_norm": 29.30799102783203, "learning_rate": 2.741510856923385e-07, "loss": 0.5142, "step": 2356 }, { "epoch": 1.7942128307633733, "grad_norm": 35.13621139526367, "learning_rate": 2.721647206074046e-07, "loss": 0.8013, "step": 2357 }, { "epoch": 1.7949743003997716, "grad_norm": 28.805770874023438, "learning_rate": 2.701853764776835e-07, "loss": 0.6767, "step": 2358 }, { "epoch": 1.7957357700361698, "grad_norm": 28.766935348510742, "learning_rate": 2.6821305624255054e-07, "loss": 0.6564, "step": 2359 }, { "epoch": 1.796497239672568, "grad_norm": 38.47027587890625, "learning_rate": 2.6624776283095343e-07, "loss": 0.751, "step": 2360 }, { "epoch": 1.7972587093089663, "grad_norm": 27.616580963134766, "learning_rate": 2.64289499161402e-07, "loss": 0.7338, "step": 2361 }, { "epoch": 1.7980201789453645, "grad_norm": 26.388782501220703, "learning_rate": 2.623382681419662e-07, "loss": 0.7544, "step": 2362 }, { "epoch": 1.798781648581763, "grad_norm": 33.01564407348633, "learning_rate": 2.603940726702747e-07, "loss": 0.6161, "step": 2363 }, { "epoch": 1.799543118218161, "grad_norm": 41.639827728271484, "learning_rate": 2.584569156335076e-07, "loss": 0.7525, "step": 2364 }, { "epoch": 1.8003045878545594, "grad_norm": 43.968692779541016, "learning_rate": 2.5652679990839145e-07, "loss": 0.3786, "step": 2365 }, { "epoch": 1.8010660574909574, "grad_norm": 37.70375442504883, "learning_rate": 2.5460372836119575e-07, "loss": 0.8499, "step": 2366 }, { "epoch": 1.8018275271273558, "grad_norm": 28.16888999938965, "learning_rate": 2.5268770384773265e-07, "loss": 0.5529, "step": 2367 }, { "epoch": 1.802588996763754, "grad_norm": 31.25115966796875, "learning_rate": 2.5077872921334623e-07, "loss": 0.6758, "step": 2368 }, { "epoch": 1.8033504664001523, "grad_norm": 28.87354278564453, "learning_rate": 2.488768072929121e-07, "loss": 0.6032, "step": 2369 }, { "epoch": 1.8041119360365505, "grad_norm": 37.87723159790039, "learning_rate": 2.4698194091083147e-07, "loss": 0.6589, "step": 2370 }, { "epoch": 1.8048734056729487, "grad_norm": 30.885589599609375, "learning_rate": 2.4509413288103135e-07, "loss": 0.7533, "step": 2371 }, { "epoch": 1.8056348753093472, "grad_norm": 29.11452865600586, "learning_rate": 2.432133860069535e-07, "loss": 0.598, "step": 2372 }, { "epoch": 1.8063963449457452, "grad_norm": 30.206979751586914, "learning_rate": 2.413397030815534e-07, "loss": 0.6931, "step": 2373 }, { "epoch": 1.8071578145821436, "grad_norm": 28.874555587768555, "learning_rate": 2.3947308688729855e-07, "loss": 0.5253, "step": 2374 }, { "epoch": 1.8079192842185416, "grad_norm": 35.28607940673828, "learning_rate": 2.3761354019616155e-07, "loss": 0.8878, "step": 2375 }, { "epoch": 1.80868075385494, "grad_norm": 37.90165328979492, "learning_rate": 2.3576106576961743e-07, "loss": 0.6989, "step": 2376 }, { "epoch": 1.8094422234913383, "grad_norm": 34.61509323120117, "learning_rate": 2.3391566635863504e-07, "loss": 0.7432, "step": 2377 }, { "epoch": 1.8102036931277365, "grad_norm": 46.2913703918457, "learning_rate": 2.3207734470368114e-07, "loss": 1.0405, "step": 2378 }, { "epoch": 1.8109651627641348, "grad_norm": 31.08005714416504, "learning_rate": 2.3024610353470978e-07, "loss": 0.6086, "step": 2379 }, { "epoch": 1.811726632400533, "grad_norm": 30.265262603759766, "learning_rate": 2.284219455711606e-07, "loss": 0.7715, "step": 2380 }, { "epoch": 1.8124881020369312, "grad_norm": 29.594951629638672, "learning_rate": 2.2660487352195337e-07, "loss": 0.7365, "step": 2381 }, { "epoch": 1.8132495716733295, "grad_norm": 23.677108764648438, "learning_rate": 2.2479489008548682e-07, "loss": 0.5344, "step": 2382 }, { "epoch": 1.814011041309728, "grad_norm": 31.178123474121094, "learning_rate": 2.229919979496331e-07, "loss": 0.6515, "step": 2383 }, { "epoch": 1.814772510946126, "grad_norm": 32.47907257080078, "learning_rate": 2.2119619979173167e-07, "loss": 0.7966, "step": 2384 }, { "epoch": 1.8155339805825244, "grad_norm": 31.520952224731445, "learning_rate": 2.1940749827858766e-07, "loss": 0.6421, "step": 2385 }, { "epoch": 1.8162954502189224, "grad_norm": 27.361095428466797, "learning_rate": 2.1762589606646955e-07, "loss": 0.5285, "step": 2386 }, { "epoch": 1.8170569198553208, "grad_norm": 24.675395965576172, "learning_rate": 2.1585139580110216e-07, "loss": 0.5731, "step": 2387 }, { "epoch": 1.817818389491719, "grad_norm": 30.639646530151367, "learning_rate": 2.1408400011766084e-07, "loss": 0.5609, "step": 2388 }, { "epoch": 1.8185798591281173, "grad_norm": 31.79228401184082, "learning_rate": 2.1232371164077558e-07, "loss": 0.7323, "step": 2389 }, { "epoch": 1.8193413287645155, "grad_norm": 27.38985824584961, "learning_rate": 2.1057053298451813e-07, "loss": 0.6571, "step": 2390 }, { "epoch": 1.8201027984009137, "grad_norm": 31.965513229370117, "learning_rate": 2.088244667524042e-07, "loss": 0.8104, "step": 2391 }, { "epoch": 1.8208642680373122, "grad_norm": 32.48710250854492, "learning_rate": 2.0708551553738409e-07, "loss": 0.6881, "step": 2392 }, { "epoch": 1.8216257376737102, "grad_norm": 31.130067825317383, "learning_rate": 2.0535368192184711e-07, "loss": 0.6157, "step": 2393 }, { "epoch": 1.8223872073101086, "grad_norm": 35.65110778808594, "learning_rate": 2.036289684776088e-07, "loss": 0.9063, "step": 2394 }, { "epoch": 1.8231486769465066, "grad_norm": 33.67428970336914, "learning_rate": 2.0191137776591263e-07, "loss": 0.5927, "step": 2395 }, { "epoch": 1.823910146582905, "grad_norm": 31.061920166015625, "learning_rate": 2.0020091233742379e-07, "loss": 0.7272, "step": 2396 }, { "epoch": 1.8246716162193033, "grad_norm": 37.102577209472656, "learning_rate": 1.9849757473222708e-07, "loss": 0.8059, "step": 2397 }, { "epoch": 1.8254330858557015, "grad_norm": 28.910085678100586, "learning_rate": 1.9680136747982248e-07, "loss": 0.6791, "step": 2398 }, { "epoch": 1.8261945554920997, "grad_norm": 28.631967544555664, "learning_rate": 1.951122930991206e-07, "loss": 0.8258, "step": 2399 }, { "epoch": 1.826956025128498, "grad_norm": 30.612184524536133, "learning_rate": 1.9343035409843946e-07, "loss": 0.6436, "step": 2400 }, { "epoch": 1.8277174947648962, "grad_norm": 29.22322654724121, "learning_rate": 1.917555529755011e-07, "loss": 0.6521, "step": 2401 }, { "epoch": 1.8284789644012944, "grad_norm": 27.466482162475586, "learning_rate": 1.9008789221742818e-07, "loss": 0.5847, "step": 2402 }, { "epoch": 1.8292404340376929, "grad_norm": 36.22343826293945, "learning_rate": 1.8842737430073754e-07, "loss": 0.7903, "step": 2403 }, { "epoch": 1.8300019036740909, "grad_norm": 30.644155502319336, "learning_rate": 1.8677400169134218e-07, "loss": 0.7817, "step": 2404 }, { "epoch": 1.8307633733104893, "grad_norm": 35.59979248046875, "learning_rate": 1.8512777684454142e-07, "loss": 0.7442, "step": 2405 }, { "epoch": 1.8315248429468873, "grad_norm": 25.276330947875977, "learning_rate": 1.8348870220502246e-07, "loss": 0.6488, "step": 2406 }, { "epoch": 1.8322863125832858, "grad_norm": 31.97323226928711, "learning_rate": 1.8185678020684937e-07, "loss": 0.8084, "step": 2407 }, { "epoch": 1.833047782219684, "grad_norm": 25.92246437072754, "learning_rate": 1.8023201327347085e-07, "loss": 0.5685, "step": 2408 }, { "epoch": 1.8338092518560822, "grad_norm": 29.91119384765625, "learning_rate": 1.7861440381770567e-07, "loss": 0.8047, "step": 2409 }, { "epoch": 1.8345707214924805, "grad_norm": 26.98511505126953, "learning_rate": 1.7700395424174456e-07, "loss": 0.5632, "step": 2410 }, { "epoch": 1.8353321911288787, "grad_norm": 34.85989761352539, "learning_rate": 1.754006669371472e-07, "loss": 0.7423, "step": 2411 }, { "epoch": 1.8360936607652771, "grad_norm": 29.09650993347168, "learning_rate": 1.7380454428483463e-07, "loss": 0.8993, "step": 2412 }, { "epoch": 1.8368551304016751, "grad_norm": 30.78036117553711, "learning_rate": 1.722155886550897e-07, "loss": 0.7101, "step": 2413 }, { "epoch": 1.8376166000380736, "grad_norm": 32.38460922241211, "learning_rate": 1.7063380240755211e-07, "loss": 0.7347, "step": 2414 }, { "epoch": 1.8383780696744716, "grad_norm": 36.91583251953125, "learning_rate": 1.69059187891214e-07, "loss": 0.7519, "step": 2415 }, { "epoch": 1.83913953931087, "grad_norm": 36.04254913330078, "learning_rate": 1.6749174744441877e-07, "loss": 0.6454, "step": 2416 }, { "epoch": 1.8399010089472683, "grad_norm": 24.678054809570312, "learning_rate": 1.6593148339485442e-07, "loss": 0.6218, "step": 2417 }, { "epoch": 1.8406624785836665, "grad_norm": 30.449003219604492, "learning_rate": 1.6437839805955315e-07, "loss": 0.631, "step": 2418 }, { "epoch": 1.8414239482200647, "grad_norm": 26.986534118652344, "learning_rate": 1.628324937448855e-07, "loss": 0.576, "step": 2419 }, { "epoch": 1.842185417856463, "grad_norm": 26.173259735107422, "learning_rate": 1.6129377274655854e-07, "loss": 0.6839, "step": 2420 }, { "epoch": 1.8429468874928612, "grad_norm": 52.379329681396484, "learning_rate": 1.5976223734961273e-07, "loss": 0.8508, "step": 2421 }, { "epoch": 1.8437083571292594, "grad_norm": 32.92402648925781, "learning_rate": 1.5823788982841592e-07, "loss": 0.8478, "step": 2422 }, { "epoch": 1.8444698267656578, "grad_norm": 31.51007652282715, "learning_rate": 1.5672073244666353e-07, "loss": 0.6684, "step": 2423 }, { "epoch": 1.8452312964020559, "grad_norm": 28.052234649658203, "learning_rate": 1.5521076745737273e-07, "loss": 0.6563, "step": 2424 }, { "epoch": 1.8459927660384543, "grad_norm": 25.36661148071289, "learning_rate": 1.537079971028793e-07, "loss": 0.4443, "step": 2425 }, { "epoch": 1.8467542356748523, "grad_norm": 26.362024307250977, "learning_rate": 1.522124236148359e-07, "loss": 0.6222, "step": 2426 }, { "epoch": 1.8475157053112508, "grad_norm": 30.097293853759766, "learning_rate": 1.5072404921420647e-07, "loss": 0.7704, "step": 2427 }, { "epoch": 1.848277174947649, "grad_norm": 26.95090675354004, "learning_rate": 1.492428761112652e-07, "loss": 0.6574, "step": 2428 }, { "epoch": 1.8490386445840472, "grad_norm": 29.468826293945312, "learning_rate": 1.4776890650559096e-07, "loss": 0.7734, "step": 2429 }, { "epoch": 1.8498001142204454, "grad_norm": 36.78318405151367, "learning_rate": 1.4630214258606723e-07, "loss": 0.8474, "step": 2430 }, { "epoch": 1.8505615838568437, "grad_norm": 27.567546844482422, "learning_rate": 1.4484258653087446e-07, "loss": 0.8402, "step": 2431 }, { "epoch": 1.851323053493242, "grad_norm": 26.077207565307617, "learning_rate": 1.433902405074905e-07, "loss": 0.5714, "step": 2432 }, { "epoch": 1.8520845231296401, "grad_norm": 26.63742446899414, "learning_rate": 1.4194510667268678e-07, "loss": 0.6973, "step": 2433 }, { "epoch": 1.8528459927660386, "grad_norm": 36.75948715209961, "learning_rate": 1.4050718717252332e-07, "loss": 0.8893, "step": 2434 }, { "epoch": 1.8536074624024366, "grad_norm": 25.275596618652344, "learning_rate": 1.3907648414234643e-07, "loss": 0.6853, "step": 2435 }, { "epoch": 1.854368932038835, "grad_norm": 30.539926528930664, "learning_rate": 1.3765299970678713e-07, "loss": 0.6317, "step": 2436 }, { "epoch": 1.8551304016752332, "grad_norm": 24.060056686401367, "learning_rate": 1.3623673597975507e-07, "loss": 0.5532, "step": 2437 }, { "epoch": 1.8558918713116315, "grad_norm": 33.06709671020508, "learning_rate": 1.3482769506443893e-07, "loss": 0.675, "step": 2438 }, { "epoch": 1.8566533409480297, "grad_norm": 32.861907958984375, "learning_rate": 1.3342587905329884e-07, "loss": 0.8757, "step": 2439 }, { "epoch": 1.857414810584428, "grad_norm": 30.79386329650879, "learning_rate": 1.3203129002806847e-07, "loss": 0.8066, "step": 2440 }, { "epoch": 1.8581762802208261, "grad_norm": 40.85968780517578, "learning_rate": 1.3064393005974674e-07, "loss": 0.8145, "step": 2441 }, { "epoch": 1.8589377498572244, "grad_norm": 35.8831672668457, "learning_rate": 1.2926380120859894e-07, "loss": 0.8362, "step": 2442 }, { "epoch": 1.8596992194936228, "grad_norm": 31.930599212646484, "learning_rate": 1.2789090552415173e-07, "loss": 0.5537, "step": 2443 }, { "epoch": 1.8604606891300208, "grad_norm": 30.79832649230957, "learning_rate": 1.2652524504518925e-07, "loss": 0.748, "step": 2444 }, { "epoch": 1.8612221587664193, "grad_norm": 30.94423484802246, "learning_rate": 1.2516682179975203e-07, "loss": 0.7341, "step": 2445 }, { "epoch": 1.8619836284028173, "grad_norm": 32.75471115112305, "learning_rate": 1.2381563780513307e-07, "loss": 0.5499, "step": 2446 }, { "epoch": 1.8627450980392157, "grad_norm": 29.783267974853516, "learning_rate": 1.224716950678756e-07, "loss": 0.66, "step": 2447 }, { "epoch": 1.863506567675614, "grad_norm": 42.51413345336914, "learning_rate": 1.2113499558376763e-07, "loss": 0.6234, "step": 2448 }, { "epoch": 1.8642680373120122, "grad_norm": 30.663572311401367, "learning_rate": 1.1980554133784183e-07, "loss": 0.5816, "step": 2449 }, { "epoch": 1.8650295069484104, "grad_norm": 27.882164001464844, "learning_rate": 1.1848333430437231e-07, "loss": 0.6182, "step": 2450 }, { "epoch": 1.8657909765848086, "grad_norm": 28.73883628845215, "learning_rate": 1.1716837644686896e-07, "loss": 0.5605, "step": 2451 }, { "epoch": 1.866552446221207, "grad_norm": 30.580446243286133, "learning_rate": 1.1586066971807864e-07, "loss": 0.5782, "step": 2452 }, { "epoch": 1.867313915857605, "grad_norm": 29.883729934692383, "learning_rate": 1.1456021605997792e-07, "loss": 0.5579, "step": 2453 }, { "epoch": 1.8680753854940035, "grad_norm": 34.98806381225586, "learning_rate": 1.1326701740377421e-07, "loss": 0.9016, "step": 2454 }, { "epoch": 1.8688368551304015, "grad_norm": 28.81910514831543, "learning_rate": 1.1198107566990024e-07, "loss": 0.8252, "step": 2455 }, { "epoch": 1.8695983247668, "grad_norm": 33.75623321533203, "learning_rate": 1.1070239276801176e-07, "loss": 0.7023, "step": 2456 }, { "epoch": 1.8703597944031982, "grad_norm": 33.16694641113281, "learning_rate": 1.0943097059698482e-07, "loss": 0.7151, "step": 2457 }, { "epoch": 1.8711212640395964, "grad_norm": 30.786767959594727, "learning_rate": 1.0816681104491467e-07, "loss": 0.6739, "step": 2458 }, { "epoch": 1.8718827336759947, "grad_norm": 35.8205680847168, "learning_rate": 1.069099159891096e-07, "loss": 0.6451, "step": 2459 }, { "epoch": 1.872644203312393, "grad_norm": 35.843631744384766, "learning_rate": 1.0566028729609046e-07, "loss": 0.7732, "step": 2460 }, { "epoch": 1.8734056729487911, "grad_norm": 32.31364059448242, "learning_rate": 1.0441792682158835e-07, "loss": 0.6867, "step": 2461 }, { "epoch": 1.8741671425851893, "grad_norm": 36.370479583740234, "learning_rate": 1.0318283641053917e-07, "loss": 0.7587, "step": 2462 }, { "epoch": 1.8749286122215878, "grad_norm": 33.773170471191406, "learning_rate": 1.0195501789708407e-07, "loss": 0.6573, "step": 2463 }, { "epoch": 1.8756900818579858, "grad_norm": 32.9802131652832, "learning_rate": 1.0073447310456453e-07, "loss": 0.6455, "step": 2464 }, { "epoch": 1.8764515514943843, "grad_norm": 29.550922393798828, "learning_rate": 9.952120384552011e-08, "loss": 0.7021, "step": 2465 }, { "epoch": 1.8772130211307823, "grad_norm": 35.218177795410156, "learning_rate": 9.83152119216868e-08, "loss": 0.8897, "step": 2466 }, { "epoch": 1.8779744907671807, "grad_norm": 30.315187454223633, "learning_rate": 9.711649912399312e-08, "loss": 0.6816, "step": 2467 }, { "epoch": 1.878735960403579, "grad_norm": 37.41974639892578, "learning_rate": 9.59250672325579e-08, "loss": 0.7715, "step": 2468 }, { "epoch": 1.8794974300399772, "grad_norm": 26.12645149230957, "learning_rate": 9.47409180166875e-08, "loss": 0.6205, "step": 2469 }, { "epoch": 1.8802588996763754, "grad_norm": 29.588632583618164, "learning_rate": 9.356405323487416e-08, "loss": 0.6226, "step": 2470 }, { "epoch": 1.8810203693127736, "grad_norm": 30.717308044433594, "learning_rate": 9.239447463479101e-08, "loss": 0.793, "step": 2471 }, { "epoch": 1.881781838949172, "grad_norm": 32.63276672363281, "learning_rate": 9.123218395329259e-08, "loss": 0.6807, "step": 2472 }, { "epoch": 1.88254330858557, "grad_norm": 27.42167854309082, "learning_rate": 9.007718291640987e-08, "loss": 0.5766, "step": 2473 }, { "epoch": 1.8833047782219685, "grad_norm": 31.356449127197266, "learning_rate": 8.892947323934864e-08, "loss": 0.7158, "step": 2474 }, { "epoch": 1.8840662478583665, "grad_norm": 28.082733154296875, "learning_rate": 8.778905662648662e-08, "loss": 0.6048, "step": 2475 }, { "epoch": 1.884827717494765, "grad_norm": 30.302377700805664, "learning_rate": 8.66559347713719e-08, "loss": 0.6184, "step": 2476 }, { "epoch": 1.8855891871311632, "grad_norm": 28.77907943725586, "learning_rate": 8.553010935671902e-08, "loss": 0.6486, "step": 2477 }, { "epoch": 1.8863506567675614, "grad_norm": 29.701671600341797, "learning_rate": 8.44115820544067e-08, "loss": 0.5627, "step": 2478 }, { "epoch": 1.8871121264039596, "grad_norm": 29.05268096923828, "learning_rate": 8.330035452547625e-08, "loss": 0.7858, "step": 2479 }, { "epoch": 1.8878735960403579, "grad_norm": 47.02728271484375, "learning_rate": 8.219642842012932e-08, "loss": 0.8236, "step": 2480 }, { "epoch": 1.888635065676756, "grad_norm": 29.845279693603516, "learning_rate": 8.109980537772455e-08, "loss": 0.6993, "step": 2481 }, { "epoch": 1.8893965353131543, "grad_norm": 30.40751075744629, "learning_rate": 8.001048702677428e-08, "loss": 0.7877, "step": 2482 }, { "epoch": 1.8901580049495528, "grad_norm": 42.87812423706055, "learning_rate": 7.892847498494394e-08, "loss": 0.8713, "step": 2483 }, { "epoch": 1.8909194745859508, "grad_norm": 27.73499298095703, "learning_rate": 7.785377085904822e-08, "loss": 0.7502, "step": 2484 }, { "epoch": 1.8916809442223492, "grad_norm": 29.514726638793945, "learning_rate": 7.678637624505213e-08, "loss": 0.6854, "step": 2485 }, { "epoch": 1.8924424138587472, "grad_norm": 27.87818717956543, "learning_rate": 7.57262927280611e-08, "loss": 0.6526, "step": 2486 }, { "epoch": 1.8932038834951457, "grad_norm": 26.554180145263672, "learning_rate": 7.467352188232745e-08, "loss": 0.6464, "step": 2487 }, { "epoch": 1.893965353131544, "grad_norm": 28.62685775756836, "learning_rate": 7.362806527124233e-08, "loss": 0.6134, "step": 2488 }, { "epoch": 1.8947268227679421, "grad_norm": 32.32271957397461, "learning_rate": 7.258992444733493e-08, "loss": 1.2043, "step": 2489 }, { "epoch": 1.8954882924043404, "grad_norm": 33.617313385009766, "learning_rate": 7.155910095227148e-08, "loss": 0.9072, "step": 2490 }, { "epoch": 1.8962497620407386, "grad_norm": 29.48932647705078, "learning_rate": 7.05355963168497e-08, "loss": 0.6956, "step": 2491 }, { "epoch": 1.897011231677137, "grad_norm": 37.154945373535156, "learning_rate": 6.95194120610021e-08, "loss": 0.8214, "step": 2492 }, { "epoch": 1.897772701313535, "grad_norm": 26.110183715820312, "learning_rate": 6.851054969378546e-08, "loss": 0.6676, "step": 2493 }, { "epoch": 1.8985341709499335, "grad_norm": 27.595172882080078, "learning_rate": 6.7509010713388e-08, "loss": 0.6528, "step": 2494 }, { "epoch": 1.8992956405863315, "grad_norm": 38.128089904785156, "learning_rate": 6.65147966071189e-08, "loss": 0.6359, "step": 2495 }, { "epoch": 1.90005711022273, "grad_norm": 34.8702392578125, "learning_rate": 6.552790885141269e-08, "loss": 0.7341, "step": 2496 }, { "epoch": 1.9008185798591282, "grad_norm": 28.02950668334961, "learning_rate": 6.454834891182204e-08, "loss": 0.7519, "step": 2497 }, { "epoch": 1.9015800494955264, "grad_norm": 34.374263763427734, "learning_rate": 6.357611824301835e-08, "loss": 0.7947, "step": 2498 }, { "epoch": 1.9023415191319246, "grad_norm": 28.59174919128418, "learning_rate": 6.261121828878835e-08, "loss": 0.5966, "step": 2499 }, { "epoch": 1.9031029887683228, "grad_norm": 31.610498428344727, "learning_rate": 6.16536504820342e-08, "loss": 0.7579, "step": 2500 }, { "epoch": 1.903864458404721, "grad_norm": 25.380033493041992, "learning_rate": 6.070341624476672e-08, "loss": 0.5611, "step": 2501 }, { "epoch": 1.9046259280411193, "grad_norm": 28.822946548461914, "learning_rate": 5.97605169881088e-08, "loss": 0.4337, "step": 2502 }, { "epoch": 1.9053873976775177, "grad_norm": 33.112823486328125, "learning_rate": 5.882495411229039e-08, "loss": 0.5238, "step": 2503 }, { "epoch": 1.9061488673139158, "grad_norm": 30.814739227294922, "learning_rate": 5.789672900664456e-08, "loss": 0.7044, "step": 2504 }, { "epoch": 1.9069103369503142, "grad_norm": 26.426315307617188, "learning_rate": 5.697584304960979e-08, "loss": 0.6399, "step": 2505 }, { "epoch": 1.9076718065867122, "grad_norm": 35.38379669189453, "learning_rate": 5.606229760872495e-08, "loss": 0.7019, "step": 2506 }, { "epoch": 1.9084332762231107, "grad_norm": 24.049610137939453, "learning_rate": 5.515609404062927e-08, "loss": 0.6042, "step": 2507 }, { "epoch": 1.9091947458595089, "grad_norm": 28.845659255981445, "learning_rate": 5.425723369105629e-08, "loss": 0.6077, "step": 2508 }, { "epoch": 1.909956215495907, "grad_norm": 36.20793151855469, "learning_rate": 5.336571789483713e-08, "loss": 0.8851, "step": 2509 }, { "epoch": 1.9107176851323053, "grad_norm": 25.311784744262695, "learning_rate": 5.2481547975895554e-08, "loss": 0.6559, "step": 2510 }, { "epoch": 1.9114791547687036, "grad_norm": 31.815767288208008, "learning_rate": 5.1604725247246243e-08, "loss": 0.6584, "step": 2511 }, { "epoch": 1.912240624405102, "grad_norm": 25.157804489135742, "learning_rate": 5.073525101099319e-08, "loss": 0.6619, "step": 2512 }, { "epoch": 1.9130020940415, "grad_norm": 24.21717071533203, "learning_rate": 4.9873126558326876e-08, "loss": 0.5912, "step": 2513 }, { "epoch": 1.9137635636778985, "grad_norm": 31.434843063354492, "learning_rate": 4.901835316952541e-08, "loss": 0.8263, "step": 2514 }, { "epoch": 1.9145250333142965, "grad_norm": 35.15153121948242, "learning_rate": 4.8170932113947855e-08, "loss": 0.8488, "step": 2515 }, { "epoch": 1.915286502950695, "grad_norm": 30.47553253173828, "learning_rate": 4.733086465003645e-08, "loss": 0.6111, "step": 2516 }, { "epoch": 1.9160479725870931, "grad_norm": 31.818838119506836, "learning_rate": 4.649815202531216e-08, "loss": 0.7565, "step": 2517 }, { "epoch": 1.9168094422234914, "grad_norm": 33.00286102294922, "learning_rate": 4.567279547637471e-08, "loss": 0.7325, "step": 2518 }, { "epoch": 1.9175709118598896, "grad_norm": 40.7569465637207, "learning_rate": 4.485479622890032e-08, "loss": 0.7324, "step": 2519 }, { "epoch": 1.9183323814962878, "grad_norm": 32.128143310546875, "learning_rate": 4.404415549763674e-08, "loss": 0.7169, "step": 2520 }, { "epoch": 1.919093851132686, "grad_norm": 48.06828689575195, "learning_rate": 4.3240874486407124e-08, "loss": 0.8055, "step": 2521 }, { "epoch": 1.9198553207690843, "grad_norm": 27.28215789794922, "learning_rate": 4.2444954388104474e-08, "loss": 0.5121, "step": 2522 }, { "epoch": 1.9206167904054827, "grad_norm": 34.6086311340332, "learning_rate": 4.1656396384689436e-08, "loss": 0.6962, "step": 2523 }, { "epoch": 1.9213782600418807, "grad_norm": 31.71796417236328, "learning_rate": 4.087520164719139e-08, "loss": 0.6104, "step": 2524 }, { "epoch": 1.9221397296782792, "grad_norm": 23.14031982421875, "learning_rate": 4.010137133570347e-08, "loss": 0.5958, "step": 2525 }, { "epoch": 1.9229011993146772, "grad_norm": 28.538768768310547, "learning_rate": 3.9334906599384215e-08, "loss": 0.6706, "step": 2526 }, { "epoch": 1.9236626689510756, "grad_norm": 37.45720672607422, "learning_rate": 3.85758085764526e-08, "loss": 0.7291, "step": 2527 }, { "epoch": 1.9244241385874739, "grad_norm": 23.553926467895508, "learning_rate": 3.782407839418856e-08, "loss": 0.6104, "step": 2528 }, { "epoch": 1.925185608223872, "grad_norm": 28.727060317993164, "learning_rate": 3.707971716893022e-08, "loss": 0.6048, "step": 2529 }, { "epoch": 1.9259470778602703, "grad_norm": 30.824804306030273, "learning_rate": 3.6342726006073916e-08, "loss": 0.8036, "step": 2530 }, { "epoch": 1.9267085474966685, "grad_norm": 28.697023391723633, "learning_rate": 3.5613106000069176e-08, "loss": 0.698, "step": 2531 }, { "epoch": 1.927470017133067, "grad_norm": 35.31669998168945, "learning_rate": 3.489085823441984e-08, "loss": 0.9521, "step": 2532 }, { "epoch": 1.928231486769465, "grad_norm": 38.04068374633789, "learning_rate": 3.417598378168352e-08, "loss": 0.9171, "step": 2533 }, { "epoch": 1.9289929564058634, "grad_norm": 32.17927932739258, "learning_rate": 3.3468483703466e-08, "loss": 0.8662, "step": 2534 }, { "epoch": 1.9297544260422614, "grad_norm": 35.836692810058594, "learning_rate": 3.2768359050422413e-08, "loss": 0.7065, "step": 2535 }, { "epoch": 1.9305158956786599, "grad_norm": 30.935548782348633, "learning_rate": 3.2075610862256634e-08, "loss": 0.854, "step": 2536 }, { "epoch": 1.9312773653150581, "grad_norm": 27.7645320892334, "learning_rate": 3.1390240167716854e-08, "loss": 0.6389, "step": 2537 }, { "epoch": 1.9320388349514563, "grad_norm": 27.734359741210938, "learning_rate": 3.07122479845956e-08, "loss": 0.6983, "step": 2538 }, { "epoch": 1.9328003045878546, "grad_norm": 35.934444427490234, "learning_rate": 3.004163531972859e-08, "loss": 0.8672, "step": 2539 }, { "epoch": 1.9335617742242528, "grad_norm": 25.42430305480957, "learning_rate": 2.9378403168993098e-08, "loss": 0.6107, "step": 2540 }, { "epoch": 1.934323243860651, "grad_norm": 31.89889907836914, "learning_rate": 2.8722552517305713e-08, "loss": 0.909, "step": 2541 }, { "epoch": 1.9350847134970492, "grad_norm": 32.79954147338867, "learning_rate": 2.8074084338620132e-08, "loss": 0.7683, "step": 2542 }, { "epoch": 1.9358461831334477, "grad_norm": 39.75404739379883, "learning_rate": 2.7432999595929376e-08, "loss": 0.9419, "step": 2543 }, { "epoch": 1.9366076527698457, "grad_norm": 28.114282608032227, "learning_rate": 2.679929924125968e-08, "loss": 0.6791, "step": 2544 }, { "epoch": 1.9373691224062441, "grad_norm": 30.516733169555664, "learning_rate": 2.6172984215672714e-08, "loss": 0.6422, "step": 2545 }, { "epoch": 1.9381305920426422, "grad_norm": 29.33420181274414, "learning_rate": 2.5554055449261705e-08, "loss": 0.8403, "step": 2546 }, { "epoch": 1.9388920616790406, "grad_norm": 31.929855346679688, "learning_rate": 2.494251386115143e-08, "loss": 0.5758, "step": 2547 }, { "epoch": 1.9396535313154388, "grad_norm": 29.950393676757812, "learning_rate": 2.4338360359497103e-08, "loss": 0.6827, "step": 2548 }, { "epoch": 1.940415000951837, "grad_norm": 25.569927215576172, "learning_rate": 2.3741595841482167e-08, "loss": 0.6004, "step": 2549 }, { "epoch": 1.9411764705882353, "grad_norm": 33.51598358154297, "learning_rate": 2.315222119331606e-08, "loss": 0.6972, "step": 2550 }, { "epoch": 1.9419379402246335, "grad_norm": 29.696245193481445, "learning_rate": 2.257023729023644e-08, "loss": 0.5973, "step": 2551 }, { "epoch": 1.942699409861032, "grad_norm": 35.83368682861328, "learning_rate": 2.1995644996503084e-08, "loss": 0.6637, "step": 2552 }, { "epoch": 1.94346087949743, "grad_norm": 30.714534759521484, "learning_rate": 2.1428445165400658e-08, "loss": 0.7483, "step": 2553 }, { "epoch": 1.9442223491338284, "grad_norm": 35.20613098144531, "learning_rate": 2.0868638639234272e-08, "loss": 0.7609, "step": 2554 }, { "epoch": 1.9449838187702264, "grad_norm": 25.279590606689453, "learning_rate": 2.031622624933227e-08, "loss": 0.5211, "step": 2555 }, { "epoch": 1.9457452884066249, "grad_norm": 33.19886016845703, "learning_rate": 1.9771208816040666e-08, "loss": 0.7672, "step": 2556 }, { "epoch": 1.946506758043023, "grad_norm": 31.1872615814209, "learning_rate": 1.92335871487237e-08, "loss": 0.8101, "step": 2557 }, { "epoch": 1.9472682276794213, "grad_norm": 26.509641647338867, "learning_rate": 1.8703362045763286e-08, "loss": 0.5709, "step": 2558 }, { "epoch": 1.9480296973158195, "grad_norm": 26.915266036987305, "learning_rate": 1.818053429455735e-08, "loss": 0.5877, "step": 2559 }, { "epoch": 1.9487911669522178, "grad_norm": 28.651647567749023, "learning_rate": 1.766510467151816e-08, "loss": 0.6055, "step": 2560 }, { "epoch": 1.949552636588616, "grad_norm": 27.7868709564209, "learning_rate": 1.7157073942071777e-08, "loss": 0.5314, "step": 2561 }, { "epoch": 1.9503141062250142, "grad_norm": 33.51142120361328, "learning_rate": 1.665644286065693e-08, "loss": 0.9414, "step": 2562 }, { "epoch": 1.9510755758614127, "grad_norm": 31.23130226135254, "learning_rate": 1.6163212170722808e-08, "loss": 0.7279, "step": 2563 }, { "epoch": 1.9518370454978107, "grad_norm": 37.194114685058594, "learning_rate": 1.567738260473073e-08, "loss": 0.8352, "step": 2564 }, { "epoch": 1.9525985151342091, "grad_norm": 27.811017990112305, "learning_rate": 1.5198954884148576e-08, "loss": 0.6223, "step": 2565 }, { "epoch": 1.9533599847706071, "grad_norm": 27.448394775390625, "learning_rate": 1.4727929719454693e-08, "loss": 0.5405, "step": 2566 }, { "epoch": 1.9541214544070056, "grad_norm": 35.92046356201172, "learning_rate": 1.4264307810132883e-08, "loss": 0.8271, "step": 2567 }, { "epoch": 1.9548829240434038, "grad_norm": 35.89884567260742, "learning_rate": 1.3808089844672967e-08, "loss": 0.8524, "step": 2568 }, { "epoch": 1.955644393679802, "grad_norm": 33.53871536254883, "learning_rate": 1.3359276500571338e-08, "loss": 0.7176, "step": 2569 }, { "epoch": 1.9564058633162003, "grad_norm": 30.455795288085938, "learning_rate": 1.2917868444325965e-08, "loss": 0.5943, "step": 2570 }, { "epoch": 1.9571673329525985, "grad_norm": 22.377416610717773, "learning_rate": 1.2483866331440275e-08, "loss": 0.5354, "step": 2571 }, { "epoch": 1.9579288025889967, "grad_norm": 35.6259765625, "learning_rate": 1.2057270806417054e-08, "loss": 0.6913, "step": 2572 }, { "epoch": 1.958690272225395, "grad_norm": 28.455875396728516, "learning_rate": 1.1638082502762326e-08, "loss": 0.7327, "step": 2573 }, { "epoch": 1.9594517418617934, "grad_norm": 29.497543334960938, "learning_rate": 1.122630204298092e-08, "loss": 0.7321, "step": 2574 }, { "epoch": 1.9602132114981914, "grad_norm": 33.515316009521484, "learning_rate": 1.082193003857812e-08, "loss": 0.6636, "step": 2575 }, { "epoch": 1.9609746811345898, "grad_norm": 35.856239318847656, "learning_rate": 1.042496709005525e-08, "loss": 0.7731, "step": 2576 }, { "epoch": 1.9617361507709878, "grad_norm": 30.422107696533203, "learning_rate": 1.0035413786913528e-08, "loss": 0.5889, "step": 2577 }, { "epoch": 1.9624976204073863, "grad_norm": 40.6138801574707, "learning_rate": 9.653270707648543e-09, "loss": 0.8379, "step": 2578 }, { "epoch": 1.9632590900437845, "grad_norm": 30.624876022338867, "learning_rate": 9.27853841975357e-09, "loss": 0.6804, "step": 2579 }, { "epoch": 1.9640205596801827, "grad_norm": 26.05486297607422, "learning_rate": 8.911217479714573e-09, "loss": 0.6214, "step": 2580 }, { "epoch": 1.964782029316581, "grad_norm": 32.497798919677734, "learning_rate": 8.55130843301244e-09, "loss": 0.5508, "step": 2581 }, { "epoch": 1.9655434989529792, "grad_norm": 26.8297119140625, "learning_rate": 8.198811814121855e-09, "loss": 0.6205, "step": 2582 }, { "epoch": 1.9663049685893776, "grad_norm": 34.03878402709961, "learning_rate": 7.853728146509088e-09, "loss": 0.7079, "step": 2583 }, { "epoch": 1.9670664382257756, "grad_norm": 30.849639892578125, "learning_rate": 7.516057942631993e-09, "loss": 0.6829, "step": 2584 }, { "epoch": 1.967827907862174, "grad_norm": 32.963623046875, "learning_rate": 7.1858017039394546e-09, "loss": 0.7279, "step": 2585 }, { "epoch": 1.968589377498572, "grad_norm": 27.354578018188477, "learning_rate": 6.862959920869716e-09, "loss": 0.6498, "step": 2586 }, { "epoch": 1.9693508471349706, "grad_norm": 29.72268295288086, "learning_rate": 6.5475330728520524e-09, "loss": 0.551, "step": 2587 }, { "epoch": 1.9701123167713688, "grad_norm": 31.215782165527344, "learning_rate": 6.23952162830288e-09, "loss": 0.5458, "step": 2588 }, { "epoch": 1.970873786407767, "grad_norm": 29.19548225402832, "learning_rate": 5.9389260446268695e-09, "loss": 0.8049, "step": 2589 }, { "epoch": 1.9716352560441652, "grad_norm": 24.525859832763672, "learning_rate": 5.645746768216387e-09, "loss": 0.6337, "step": 2590 }, { "epoch": 1.9723967256805635, "grad_norm": 24.5540714263916, "learning_rate": 5.359984234449278e-09, "loss": 0.7272, "step": 2591 }, { "epoch": 1.9731581953169617, "grad_norm": 34.44743347167969, "learning_rate": 5.081638867691086e-09, "loss": 0.7913, "step": 2592 }, { "epoch": 1.97391966495336, "grad_norm": 25.064788818359375, "learning_rate": 4.810711081291719e-09, "loss": 0.6335, "step": 2593 }, { "epoch": 1.9746811345897584, "grad_norm": 31.73797035217285, "learning_rate": 4.547201277585456e-09, "loss": 0.6793, "step": 2594 }, { "epoch": 1.9754426042261564, "grad_norm": 37.75809860229492, "learning_rate": 4.291109847890385e-09, "loss": 0.6341, "step": 2595 }, { "epoch": 1.9762040738625548, "grad_norm": 29.767826080322266, "learning_rate": 4.042437172510072e-09, "loss": 0.6231, "step": 2596 }, { "epoch": 1.9769655434989528, "grad_norm": 28.760480880737305, "learning_rate": 3.801183620729121e-09, "loss": 0.7008, "step": 2597 }, { "epoch": 1.9777270131353513, "grad_norm": 32.87038040161133, "learning_rate": 3.5673495508148357e-09, "loss": 0.7481, "step": 2598 }, { "epoch": 1.9784884827717495, "grad_norm": 27.856800079345703, "learning_rate": 3.340935310017224e-09, "loss": 0.6696, "step": 2599 }, { "epoch": 1.9792499524081477, "grad_norm": 42.673458099365234, "learning_rate": 3.1219412345673274e-09, "loss": 0.7877, "step": 2600 }, { "epoch": 1.980011422044546, "grad_norm": 37.44389724731445, "learning_rate": 2.9103676496761156e-09, "loss": 0.866, "step": 2601 }, { "epoch": 1.9807728916809442, "grad_norm": 31.068260192871094, "learning_rate": 2.706214869536705e-09, "loss": 0.581, "step": 2602 }, { "epoch": 1.9815343613173426, "grad_norm": 36.26176452636719, "learning_rate": 2.5094831973199175e-09, "loss": 0.7472, "step": 2603 }, { "epoch": 1.9822958309537406, "grad_norm": 27.535024642944336, "learning_rate": 2.320172925178721e-09, "loss": 0.7681, "step": 2604 }, { "epoch": 1.983057300590139, "grad_norm": 30.589597702026367, "learning_rate": 2.1382843342426795e-09, "loss": 0.6945, "step": 2605 }, { "epoch": 1.983818770226537, "grad_norm": 28.598867416381836, "learning_rate": 1.9638176946212837e-09, "loss": 0.5873, "step": 2606 }, { "epoch": 1.9845802398629355, "grad_norm": 30.991840362548828, "learning_rate": 1.79677326540173e-09, "loss": 0.6344, "step": 2607 }, { "epoch": 1.9853417094993338, "grad_norm": 44.96479034423828, "learning_rate": 1.6371512946489199e-09, "loss": 0.8516, "step": 2608 }, { "epoch": 1.986103179135732, "grad_norm": 28.062541961669922, "learning_rate": 1.4849520194065714e-09, "loss": 0.6928, "step": 2609 }, { "epoch": 1.9868646487721302, "grad_norm": 36.74098587036133, "learning_rate": 1.340175665693888e-09, "loss": 0.7047, "step": 2610 }, { "epoch": 1.9876261184085284, "grad_norm": 36.37099075317383, "learning_rate": 1.2028224485072237e-09, "loss": 0.8266, "step": 2611 }, { "epoch": 1.9883875880449267, "grad_norm": 24.739198684692383, "learning_rate": 1.0728925718195282e-09, "loss": 0.5258, "step": 2612 }, { "epoch": 1.9891490576813249, "grad_norm": 28.954843521118164, "learning_rate": 9.503862285803467e-10, "loss": 0.6344, "step": 2613 }, { "epoch": 1.9899105273177233, "grad_norm": 31.088581085205078, "learning_rate": 8.353036007147097e-10, "loss": 0.8029, "step": 2614 }, { "epoch": 1.9906719969541213, "grad_norm": 47.386165618896484, "learning_rate": 7.276448591231334e-10, "loss": 0.9589, "step": 2615 }, { "epoch": 1.9914334665905198, "grad_norm": 32.6035041809082, "learning_rate": 6.274101636816188e-10, "loss": 0.8384, "step": 2616 }, { "epoch": 1.9921949362269178, "grad_norm": 33.007789611816406, "learning_rate": 5.345996632405425e-10, "loss": 0.6737, "step": 2617 }, { "epoch": 1.9929564058633162, "grad_norm": 25.595951080322266, "learning_rate": 4.4921349562687676e-10, "loss": 0.6961, "step": 2618 }, { "epoch": 1.9937178754997145, "grad_norm": 28.480968475341797, "learning_rate": 3.7125178764085835e-10, "loss": 0.6405, "step": 2619 }, { "epoch": 1.9944793451361127, "grad_norm": 32.006690979003906, "learning_rate": 3.007146550570994e-10, "loss": 0.768, "step": 2620 }, { "epoch": 1.995240814772511, "grad_norm": 35.4099006652832, "learning_rate": 2.3760220262569746e-10, "loss": 0.717, "step": 2621 }, { "epoch": 1.9960022844089091, "grad_norm": 24.73089027404785, "learning_rate": 1.8191452407001486e-10, "loss": 0.5761, "step": 2622 }, { "epoch": 1.9967637540453076, "grad_norm": 25.563087463378906, "learning_rate": 1.336517020877892e-10, "loss": 0.59, "step": 2623 }, { "epoch": 1.9975252236817056, "grad_norm": 28.94989013671875, "learning_rate": 9.281380835057807e-11, "loss": 0.5689, "step": 2624 }, { "epoch": 1.998286693318104, "grad_norm": 27.524860382080078, "learning_rate": 5.940090350320393e-11, "loss": 0.6398, "step": 2625 }, { "epoch": 1.999048162954502, "grad_norm": 41.85504150390625, "learning_rate": 3.3413037164864437e-11, "loss": 1.4401, "step": 2626 }, { "epoch": 1.9998096325909005, "grad_norm": 23.15554428100586, "learning_rate": 1.4850247928577255e-11, "loss": 0.5939, "step": 2627 }, { "epoch": 2.0, "grad_norm": 42.80801010131836, "learning_rate": 3.712563360624977e-12, "loss": 0.4714, "step": 2628 }, { "epoch": 2.0007614696363984, "grad_norm": 32.09111022949219, "learning_rate": 0.0, "loss": 0.6621, "step": 2629 }, { "epoch": 2.0015229392727965, "grad_norm": 36.724910736083984, "learning_rate": 2.554957285044737e-06, "loss": 0.809, "step": 2630 }, { "epoch": 2.002284408909195, "grad_norm": 29.856298446655273, "learning_rate": 2.5514375972831057e-06, "loss": 0.6297, "step": 2631 }, { "epoch": 2.003045878545593, "grad_norm": 29.63297462463379, "learning_rate": 2.547919504905488e-06, "loss": 0.6814, "step": 2632 }, { "epoch": 2.0038073481819914, "grad_norm": 34.38152313232422, "learning_rate": 2.5444030102041263e-06, "loss": 0.6104, "step": 2633 }, { "epoch": 2.0045688178183894, "grad_norm": 25.15961265563965, "learning_rate": 2.540888115470226e-06, "loss": 0.6449, "step": 2634 }, { "epoch": 2.005330287454788, "grad_norm": 36.988372802734375, "learning_rate": 2.5373748229939533e-06, "loss": 0.9041, "step": 2635 }, { "epoch": 2.006091757091186, "grad_norm": 38.70914840698242, "learning_rate": 2.533863135064425e-06, "loss": 0.7159, "step": 2636 }, { "epoch": 2.0068532267275843, "grad_norm": 32.377037048339844, "learning_rate": 2.530353053969716e-06, "loss": 0.6373, "step": 2637 }, { "epoch": 2.0076146963639823, "grad_norm": 27.70588493347168, "learning_rate": 2.526844581996852e-06, "loss": 0.5263, "step": 2638 }, { "epoch": 2.0083761660003807, "grad_norm": 33.944122314453125, "learning_rate": 2.523337721431809e-06, "loss": 0.7145, "step": 2639 }, { "epoch": 2.009137635636779, "grad_norm": 31.66524314880371, "learning_rate": 2.5198324745595214e-06, "loss": 0.7393, "step": 2640 }, { "epoch": 2.009899105273177, "grad_norm": 33.91412353515625, "learning_rate": 2.51632884366386e-06, "loss": 0.7499, "step": 2641 }, { "epoch": 2.0106605749095756, "grad_norm": 28.268230438232422, "learning_rate": 2.5128268310276505e-06, "loss": 0.7171, "step": 2642 }, { "epoch": 2.0114220445459736, "grad_norm": 28.64215087890625, "learning_rate": 2.509326438932661e-06, "loss": 0.6073, "step": 2643 }, { "epoch": 2.012183514182372, "grad_norm": 32.430320739746094, "learning_rate": 2.5058276696596084e-06, "loss": 0.7494, "step": 2644 }, { "epoch": 2.01294498381877, "grad_norm": 35.946231842041016, "learning_rate": 2.502330525488147e-06, "loss": 0.7623, "step": 2645 }, { "epoch": 2.0137064534551685, "grad_norm": 30.655607223510742, "learning_rate": 2.4988350086968744e-06, "loss": 0.665, "step": 2646 }, { "epoch": 2.0144679230915665, "grad_norm": 34.46199417114258, "learning_rate": 2.4953411215633278e-06, "loss": 0.8534, "step": 2647 }, { "epoch": 2.015229392727965, "grad_norm": 34.27285385131836, "learning_rate": 2.4918488663639824e-06, "loss": 0.7143, "step": 2648 }, { "epoch": 2.0159908623643634, "grad_norm": 29.49825096130371, "learning_rate": 2.4883582453742495e-06, "loss": 0.6247, "step": 2649 }, { "epoch": 2.0167523320007614, "grad_norm": 25.18963050842285, "learning_rate": 2.4848692608684773e-06, "loss": 0.6644, "step": 2650 }, { "epoch": 2.01751380163716, "grad_norm": 31.328371047973633, "learning_rate": 2.4813819151199435e-06, "loss": 0.7041, "step": 2651 }, { "epoch": 2.018275271273558, "grad_norm": 27.792232513427734, "learning_rate": 2.4778962104008664e-06, "loss": 0.6675, "step": 2652 }, { "epoch": 2.0190367409099563, "grad_norm": 33.84516906738281, "learning_rate": 2.474412148982387e-06, "loss": 0.635, "step": 2653 }, { "epoch": 2.0197982105463543, "grad_norm": 33.45273971557617, "learning_rate": 2.470929733134579e-06, "loss": 0.7197, "step": 2654 }, { "epoch": 2.020559680182753, "grad_norm": 30.420703887939453, "learning_rate": 2.4674489651264433e-06, "loss": 0.672, "step": 2655 }, { "epoch": 2.021321149819151, "grad_norm": 36.99637222290039, "learning_rate": 2.463969847225907e-06, "loss": 0.7627, "step": 2656 }, { "epoch": 2.0220826194555492, "grad_norm": 36.0267448425293, "learning_rate": 2.4604923816998226e-06, "loss": 0.7432, "step": 2657 }, { "epoch": 2.0228440890919472, "grad_norm": 30.906543731689453, "learning_rate": 2.4570165708139653e-06, "loss": 0.7139, "step": 2658 }, { "epoch": 2.0236055587283457, "grad_norm": 35.617088317871094, "learning_rate": 2.4535424168330336e-06, "loss": 0.7678, "step": 2659 }, { "epoch": 2.024367028364744, "grad_norm": 30.224918365478516, "learning_rate": 2.4500699220206418e-06, "loss": 0.5453, "step": 2660 }, { "epoch": 2.025128498001142, "grad_norm": 33.075870513916016, "learning_rate": 2.446599088639332e-06, "loss": 0.576, "step": 2661 }, { "epoch": 2.0258899676375406, "grad_norm": 29.24921989440918, "learning_rate": 2.4431299189505563e-06, "loss": 0.6705, "step": 2662 }, { "epoch": 2.0266514372739386, "grad_norm": 36.619667053222656, "learning_rate": 2.4396624152146854e-06, "loss": 0.6387, "step": 2663 }, { "epoch": 2.027412906910337, "grad_norm": 27.41086769104004, "learning_rate": 2.436196579691003e-06, "loss": 0.6311, "step": 2664 }, { "epoch": 2.028174376546735, "grad_norm": 31.923843383789062, "learning_rate": 2.432732414637708e-06, "loss": 0.708, "step": 2665 }, { "epoch": 2.0289358461831335, "grad_norm": 28.00581169128418, "learning_rate": 2.42926992231191e-06, "loss": 0.6332, "step": 2666 }, { "epoch": 2.0296973158195315, "grad_norm": 37.546512603759766, "learning_rate": 2.4258091049696275e-06, "loss": 0.6442, "step": 2667 }, { "epoch": 2.03045878545593, "grad_norm": 37.19662094116211, "learning_rate": 2.4223499648657884e-06, "loss": 0.7178, "step": 2668 }, { "epoch": 2.0312202550923284, "grad_norm": 27.53927230834961, "learning_rate": 2.418892504254231e-06, "loss": 0.4716, "step": 2669 }, { "epoch": 2.0319817247287264, "grad_norm": 33.06608200073242, "learning_rate": 2.4154367253876954e-06, "loss": 0.6845, "step": 2670 }, { "epoch": 2.032743194365125, "grad_norm": 30.894197463989258, "learning_rate": 2.411982630517828e-06, "loss": 0.6381, "step": 2671 }, { "epoch": 2.033504664001523, "grad_norm": 28.420122146606445, "learning_rate": 2.408530221895172e-06, "loss": 0.7431, "step": 2672 }, { "epoch": 2.0342661336379213, "grad_norm": 54.150962829589844, "learning_rate": 2.4050795017691823e-06, "loss": 0.7031, "step": 2673 }, { "epoch": 2.0350276032743193, "grad_norm": 33.161766052246094, "learning_rate": 2.4016304723882062e-06, "loss": 0.8263, "step": 2674 }, { "epoch": 2.0357890729107178, "grad_norm": 37.252525329589844, "learning_rate": 2.398183135999492e-06, "loss": 0.9176, "step": 2675 }, { "epoch": 2.0365505425471158, "grad_norm": 32.252681732177734, "learning_rate": 2.394737494849184e-06, "loss": 0.6592, "step": 2676 }, { "epoch": 2.037312012183514, "grad_norm": 34.02185821533203, "learning_rate": 2.3912935511823203e-06, "loss": 0.7618, "step": 2677 }, { "epoch": 2.038073481819912, "grad_norm": 31.590497970581055, "learning_rate": 2.387851307242839e-06, "loss": 0.6062, "step": 2678 }, { "epoch": 2.0388349514563107, "grad_norm": 34.79767990112305, "learning_rate": 2.3844107652735666e-06, "loss": 0.6851, "step": 2679 }, { "epoch": 2.039596421092709, "grad_norm": 38.405025482177734, "learning_rate": 2.3809719275162154e-06, "loss": 0.7276, "step": 2680 }, { "epoch": 2.040357890729107, "grad_norm": 33.83439254760742, "learning_rate": 2.3775347962113975e-06, "loss": 0.7195, "step": 2681 }, { "epoch": 2.0411193603655056, "grad_norm": 28.758760452270508, "learning_rate": 2.374099373598608e-06, "loss": 0.6955, "step": 2682 }, { "epoch": 2.0418808300019036, "grad_norm": 37.01749801635742, "learning_rate": 2.3706656619162278e-06, "loss": 0.6323, "step": 2683 }, { "epoch": 2.042642299638302, "grad_norm": 34.21178436279297, "learning_rate": 2.3672336634015243e-06, "loss": 0.6263, "step": 2684 }, { "epoch": 2.0434037692747, "grad_norm": 25.095067977905273, "learning_rate": 2.3638033802906468e-06, "loss": 0.4878, "step": 2685 }, { "epoch": 2.0441652389110985, "grad_norm": 30.67310333251953, "learning_rate": 2.3603748148186347e-06, "loss": 0.541, "step": 2686 }, { "epoch": 2.0449267085474965, "grad_norm": 39.88631057739258, "learning_rate": 2.3569479692193964e-06, "loss": 0.7745, "step": 2687 }, { "epoch": 2.045688178183895, "grad_norm": 28.737600326538086, "learning_rate": 2.3535228457257283e-06, "loss": 0.5135, "step": 2688 }, { "epoch": 2.0464496478202934, "grad_norm": 32.8947868347168, "learning_rate": 2.350099446569299e-06, "loss": 0.7244, "step": 2689 }, { "epoch": 2.0472111174566914, "grad_norm": 28.975154876708984, "learning_rate": 2.34667777398066e-06, "loss": 0.528, "step": 2690 }, { "epoch": 2.04797258709309, "grad_norm": 35.946598052978516, "learning_rate": 2.3432578301892346e-06, "loss": 0.7916, "step": 2691 }, { "epoch": 2.048734056729488, "grad_norm": 26.219942092895508, "learning_rate": 2.339839617423318e-06, "loss": 0.6149, "step": 2692 }, { "epoch": 2.0494955263658863, "grad_norm": 38.671321868896484, "learning_rate": 2.3364231379100786e-06, "loss": 0.7621, "step": 2693 }, { "epoch": 2.0502569960022843, "grad_norm": 37.21269607543945, "learning_rate": 2.333008393875558e-06, "loss": 0.8555, "step": 2694 }, { "epoch": 2.0510184656386827, "grad_norm": 33.47099685668945, "learning_rate": 2.329595387544663e-06, "loss": 0.6881, "step": 2695 }, { "epoch": 2.0517799352750807, "grad_norm": 38.447967529296875, "learning_rate": 2.3261841211411718e-06, "loss": 0.7767, "step": 2696 }, { "epoch": 2.052541404911479, "grad_norm": 28.296552658081055, "learning_rate": 2.322774596887726e-06, "loss": 0.5923, "step": 2697 }, { "epoch": 2.053302874547877, "grad_norm": 32.68894958496094, "learning_rate": 2.3193668170058314e-06, "loss": 0.6151, "step": 2698 }, { "epoch": 2.0540643441842756, "grad_norm": 33.723575592041016, "learning_rate": 2.3159607837158645e-06, "loss": 0.6879, "step": 2699 }, { "epoch": 2.054825813820674, "grad_norm": 30.567623138427734, "learning_rate": 2.3125564992370553e-06, "loss": 0.7506, "step": 2700 }, { "epoch": 2.055587283457072, "grad_norm": 44.19010925292969, "learning_rate": 2.3091539657874984e-06, "loss": 0.9602, "step": 2701 }, { "epoch": 2.0563487530934705, "grad_norm": 26.454425811767578, "learning_rate": 2.3057531855841464e-06, "loss": 0.6076, "step": 2702 }, { "epoch": 2.0571102227298685, "grad_norm": 35.36713409423828, "learning_rate": 2.302354160842809e-06, "loss": 0.7093, "step": 2703 }, { "epoch": 2.057871692366267, "grad_norm": 27.136823654174805, "learning_rate": 2.298956893778154e-06, "loss": 0.6123, "step": 2704 }, { "epoch": 2.058633162002665, "grad_norm": 32.624664306640625, "learning_rate": 2.2955613866037037e-06, "loss": 0.6473, "step": 2705 }, { "epoch": 2.0593946316390634, "grad_norm": 28.906478881835938, "learning_rate": 2.2921676415318288e-06, "loss": 0.4768, "step": 2706 }, { "epoch": 2.0601561012754614, "grad_norm": 37.861358642578125, "learning_rate": 2.288775660773761e-06, "loss": 0.7586, "step": 2707 }, { "epoch": 2.06091757091186, "grad_norm": 32.612815856933594, "learning_rate": 2.2853854465395758e-06, "loss": 0.7224, "step": 2708 }, { "epoch": 2.0616790405482583, "grad_norm": 35.3125, "learning_rate": 2.281997001038199e-06, "loss": 0.6323, "step": 2709 }, { "epoch": 2.0624405101846564, "grad_norm": 29.31070899963379, "learning_rate": 2.278610326477405e-06, "loss": 0.5617, "step": 2710 }, { "epoch": 2.063201979821055, "grad_norm": 35.75231170654297, "learning_rate": 2.275225425063813e-06, "loss": 0.7882, "step": 2711 }, { "epoch": 2.063963449457453, "grad_norm": 22.62790298461914, "learning_rate": 2.271842299002887e-06, "loss": 0.4202, "step": 2712 }, { "epoch": 2.0647249190938513, "grad_norm": 27.09946632385254, "learning_rate": 2.268460950498936e-06, "loss": 0.6382, "step": 2713 }, { "epoch": 2.0654863887302493, "grad_norm": 30.387042999267578, "learning_rate": 2.2650813817551094e-06, "loss": 0.5457, "step": 2714 }, { "epoch": 2.0662478583666477, "grad_norm": 31.93035125732422, "learning_rate": 2.2617035949733946e-06, "loss": 0.559, "step": 2715 }, { "epoch": 2.0670093280030457, "grad_norm": 30.75739860534668, "learning_rate": 2.258327592354625e-06, "loss": 0.6859, "step": 2716 }, { "epoch": 2.067770797639444, "grad_norm": 30.511552810668945, "learning_rate": 2.2549533760984645e-06, "loss": 0.5363, "step": 2717 }, { "epoch": 2.068532267275842, "grad_norm": 30.775714874267578, "learning_rate": 2.251580948403416e-06, "loss": 0.7123, "step": 2718 }, { "epoch": 2.0692937369122406, "grad_norm": 35.703521728515625, "learning_rate": 2.2482103114668174e-06, "loss": 0.6828, "step": 2719 }, { "epoch": 2.070055206548639, "grad_norm": 27.549148559570312, "learning_rate": 2.2448414674848385e-06, "loss": 0.6421, "step": 2720 }, { "epoch": 2.070816676185037, "grad_norm": 27.492868423461914, "learning_rate": 2.241474418652481e-06, "loss": 0.6515, "step": 2721 }, { "epoch": 2.0715781458214355, "grad_norm": 37.95566177368164, "learning_rate": 2.2381091671635787e-06, "loss": 0.8889, "step": 2722 }, { "epoch": 2.0723396154578335, "grad_norm": 26.991741180419922, "learning_rate": 2.23474571521079e-06, "loss": 0.7156, "step": 2723 }, { "epoch": 2.073101085094232, "grad_norm": 32.10348129272461, "learning_rate": 2.2313840649856077e-06, "loss": 0.647, "step": 2724 }, { "epoch": 2.07386255473063, "grad_norm": 28.42721176147461, "learning_rate": 2.2280242186783473e-06, "loss": 0.6415, "step": 2725 }, { "epoch": 2.0746240243670284, "grad_norm": 31.033693313598633, "learning_rate": 2.224666178478144e-06, "loss": 0.5502, "step": 2726 }, { "epoch": 2.0753854940034264, "grad_norm": 43.40159225463867, "learning_rate": 2.221309946572962e-06, "loss": 0.5775, "step": 2727 }, { "epoch": 2.076146963639825, "grad_norm": 34.83430480957031, "learning_rate": 2.2179555251495887e-06, "loss": 0.758, "step": 2728 }, { "epoch": 2.0769084332762233, "grad_norm": 28.433305740356445, "learning_rate": 2.214602916393628e-06, "loss": 0.5911, "step": 2729 }, { "epoch": 2.0776699029126213, "grad_norm": 34.25566101074219, "learning_rate": 2.211252122489502e-06, "loss": 0.6403, "step": 2730 }, { "epoch": 2.0784313725490198, "grad_norm": 27.0931396484375, "learning_rate": 2.2079031456204545e-06, "loss": 0.6823, "step": 2731 }, { "epoch": 2.0791928421854178, "grad_norm": 26.47356414794922, "learning_rate": 2.204555987968541e-06, "loss": 0.6979, "step": 2732 }, { "epoch": 2.0799543118218162, "grad_norm": 29.564836502075195, "learning_rate": 2.201210651714637e-06, "loss": 0.5736, "step": 2733 }, { "epoch": 2.0807157814582142, "grad_norm": 40.11194610595703, "learning_rate": 2.197867139038425e-06, "loss": 0.8792, "step": 2734 }, { "epoch": 2.0814772510946127, "grad_norm": 35.69405746459961, "learning_rate": 2.194525452118403e-06, "loss": 0.6898, "step": 2735 }, { "epoch": 2.0822387207310107, "grad_norm": 34.791908264160156, "learning_rate": 2.191185593131877e-06, "loss": 0.7469, "step": 2736 }, { "epoch": 2.083000190367409, "grad_norm": 42.962711334228516, "learning_rate": 2.1878475642549677e-06, "loss": 0.8187, "step": 2737 }, { "epoch": 2.083761660003807, "grad_norm": 36.920936584472656, "learning_rate": 2.1845113676625973e-06, "loss": 0.7682, "step": 2738 }, { "epoch": 2.0845231296402056, "grad_norm": 32.207969665527344, "learning_rate": 2.1811770055284968e-06, "loss": 0.805, "step": 2739 }, { "epoch": 2.085284599276604, "grad_norm": 36.39158630371094, "learning_rate": 2.1778444800251985e-06, "loss": 0.8209, "step": 2740 }, { "epoch": 2.086046068913002, "grad_norm": 38.26339340209961, "learning_rate": 2.1745137933240478e-06, "loss": 0.707, "step": 2741 }, { "epoch": 2.0868075385494005, "grad_norm": 33.83285140991211, "learning_rate": 2.1711849475951785e-06, "loss": 0.5886, "step": 2742 }, { "epoch": 2.0875690081857985, "grad_norm": 26.331220626831055, "learning_rate": 2.1678579450075338e-06, "loss": 0.4485, "step": 2743 }, { "epoch": 2.088330477822197, "grad_norm": 26.193998336791992, "learning_rate": 2.1645327877288518e-06, "loss": 0.4992, "step": 2744 }, { "epoch": 2.089091947458595, "grad_norm": 29.64794921875, "learning_rate": 2.161209477925674e-06, "loss": 0.6692, "step": 2745 }, { "epoch": 2.0898534170949934, "grad_norm": 38.8067626953125, "learning_rate": 2.157888017763333e-06, "loss": 0.8341, "step": 2746 }, { "epoch": 2.0906148867313914, "grad_norm": 33.91508483886719, "learning_rate": 2.1545684094059572e-06, "loss": 0.6789, "step": 2747 }, { "epoch": 2.09137635636779, "grad_norm": 35.92626190185547, "learning_rate": 2.1512506550164687e-06, "loss": 0.8419, "step": 2748 }, { "epoch": 2.0921378260041883, "grad_norm": 33.8650016784668, "learning_rate": 2.147934756756581e-06, "loss": 0.5864, "step": 2749 }, { "epoch": 2.0928992956405863, "grad_norm": 29.450340270996094, "learning_rate": 2.1446207167868004e-06, "loss": 0.7212, "step": 2750 }, { "epoch": 2.0936607652769847, "grad_norm": 30.898523330688477, "learning_rate": 2.1413085372664203e-06, "loss": 0.716, "step": 2751 }, { "epoch": 2.0944222349133828, "grad_norm": 33.348514556884766, "learning_rate": 2.137998220353523e-06, "loss": 0.7381, "step": 2752 }, { "epoch": 2.095183704549781, "grad_norm": 35.20305633544922, "learning_rate": 2.134689768204975e-06, "loss": 0.601, "step": 2753 }, { "epoch": 2.095945174186179, "grad_norm": 30.802499771118164, "learning_rate": 2.131383182976433e-06, "loss": 0.7411, "step": 2754 }, { "epoch": 2.0967066438225777, "grad_norm": 33.5802001953125, "learning_rate": 2.128078466822332e-06, "loss": 0.7578, "step": 2755 }, { "epoch": 2.0974681134589757, "grad_norm": 32.02031707763672, "learning_rate": 2.124775621895891e-06, "loss": 0.6319, "step": 2756 }, { "epoch": 2.098229583095374, "grad_norm": 29.34837532043457, "learning_rate": 2.1214746503491102e-06, "loss": 0.6798, "step": 2757 }, { "epoch": 2.098991052731772, "grad_norm": 31.122468948364258, "learning_rate": 2.1181755543327687e-06, "loss": 0.6357, "step": 2758 }, { "epoch": 2.0997525223681706, "grad_norm": 36.041221618652344, "learning_rate": 2.114878335996424e-06, "loss": 0.6243, "step": 2759 }, { "epoch": 2.100513992004569, "grad_norm": 33.60100555419922, "learning_rate": 2.1115829974884097e-06, "loss": 0.8179, "step": 2760 }, { "epoch": 2.101275461640967, "grad_norm": 34.87454605102539, "learning_rate": 2.108289540955832e-06, "loss": 0.7477, "step": 2761 }, { "epoch": 2.1020369312773655, "grad_norm": 36.94026565551758, "learning_rate": 2.104997968544579e-06, "loss": 0.7735, "step": 2762 }, { "epoch": 2.1027984009137635, "grad_norm": 30.245534896850586, "learning_rate": 2.101708282399304e-06, "loss": 0.584, "step": 2763 }, { "epoch": 2.103559870550162, "grad_norm": 28.392004013061523, "learning_rate": 2.0984204846634338e-06, "loss": 0.6877, "step": 2764 }, { "epoch": 2.10432134018656, "grad_norm": 28.813669204711914, "learning_rate": 2.0951345774791594e-06, "loss": 0.7731, "step": 2765 }, { "epoch": 2.1050828098229584, "grad_norm": 29.806596755981445, "learning_rate": 2.0918505629874493e-06, "loss": 0.651, "step": 2766 }, { "epoch": 2.1058442794593564, "grad_norm": 29.53590202331543, "learning_rate": 2.0885684433280336e-06, "loss": 0.6625, "step": 2767 }, { "epoch": 2.106605749095755, "grad_norm": 28.497465133666992, "learning_rate": 2.0852882206394076e-06, "loss": 0.6468, "step": 2768 }, { "epoch": 2.1073672187321533, "grad_norm": 37.14005661010742, "learning_rate": 2.082009897058832e-06, "loss": 0.6511, "step": 2769 }, { "epoch": 2.1081286883685513, "grad_norm": 40.95936584472656, "learning_rate": 2.0787334747223274e-06, "loss": 0.7842, "step": 2770 }, { "epoch": 2.1088901580049497, "grad_norm": 34.1695442199707, "learning_rate": 2.075458955764682e-06, "loss": 0.6759, "step": 2771 }, { "epoch": 2.1096516276413477, "grad_norm": 33.675811767578125, "learning_rate": 2.0721863423194386e-06, "loss": 0.5501, "step": 2772 }, { "epoch": 2.110413097277746, "grad_norm": 37.50621795654297, "learning_rate": 2.068915636518897e-06, "loss": 0.759, "step": 2773 }, { "epoch": 2.111174566914144, "grad_norm": 39.66071319580078, "learning_rate": 2.065646840494115e-06, "loss": 0.7729, "step": 2774 }, { "epoch": 2.1119360365505426, "grad_norm": 34.54794692993164, "learning_rate": 2.0623799563749125e-06, "loss": 0.822, "step": 2775 }, { "epoch": 2.1126975061869406, "grad_norm": 45.4400749206543, "learning_rate": 2.0591149862898566e-06, "loss": 0.7598, "step": 2776 }, { "epoch": 2.113458975823339, "grad_norm": 29.210983276367188, "learning_rate": 2.05585193236627e-06, "loss": 0.6385, "step": 2777 }, { "epoch": 2.114220445459737, "grad_norm": 29.053844451904297, "learning_rate": 2.0525907967302234e-06, "loss": 0.6719, "step": 2778 }, { "epoch": 2.1149819150961355, "grad_norm": 29.968189239501953, "learning_rate": 2.0493315815065453e-06, "loss": 0.7329, "step": 2779 }, { "epoch": 2.115743384732534, "grad_norm": 32.46784210205078, "learning_rate": 2.0460742888188074e-06, "loss": 0.6636, "step": 2780 }, { "epoch": 2.116504854368932, "grad_norm": 36.47773361206055, "learning_rate": 2.042818920789326e-06, "loss": 0.6831, "step": 2781 }, { "epoch": 2.1172663240053304, "grad_norm": 34.53937911987305, "learning_rate": 2.039565479539168e-06, "loss": 0.6681, "step": 2782 }, { "epoch": 2.1180277936417284, "grad_norm": 23.8504581451416, "learning_rate": 2.036313967188146e-06, "loss": 0.656, "step": 2783 }, { "epoch": 2.118789263278127, "grad_norm": 34.38410568237305, "learning_rate": 2.0330643858548133e-06, "loss": 0.6077, "step": 2784 }, { "epoch": 2.119550732914525, "grad_norm": 29.759443283081055, "learning_rate": 2.029816737656464e-06, "loss": 0.6074, "step": 2785 }, { "epoch": 2.1203122025509233, "grad_norm": 24.493896484375, "learning_rate": 2.0265710247091353e-06, "loss": 0.5803, "step": 2786 }, { "epoch": 2.1210736721873213, "grad_norm": 38.59502410888672, "learning_rate": 2.0233272491275994e-06, "loss": 0.9125, "step": 2787 }, { "epoch": 2.12183514182372, "grad_norm": 34.60749816894531, "learning_rate": 2.020085413025375e-06, "loss": 0.5254, "step": 2788 }, { "epoch": 2.122596611460118, "grad_norm": 28.385862350463867, "learning_rate": 2.016845518514705e-06, "loss": 0.7274, "step": 2789 }, { "epoch": 2.1233580810965162, "grad_norm": 29.854501724243164, "learning_rate": 2.0136075677065764e-06, "loss": 0.6945, "step": 2790 }, { "epoch": 2.1241195507329147, "grad_norm": 31.9979305267334, "learning_rate": 2.0103715627107032e-06, "loss": 0.7494, "step": 2791 }, { "epoch": 2.1248810203693127, "grad_norm": 36.22807312011719, "learning_rate": 2.00713750563554e-06, "loss": 0.6436, "step": 2792 }, { "epoch": 2.125642490005711, "grad_norm": 32.35902786254883, "learning_rate": 2.003905398588265e-06, "loss": 0.5797, "step": 2793 }, { "epoch": 2.126403959642109, "grad_norm": 38.77256774902344, "learning_rate": 2.000675243674788e-06, "loss": 0.7746, "step": 2794 }, { "epoch": 2.1271654292785076, "grad_norm": 25.81340980529785, "learning_rate": 1.9974470429997482e-06, "loss": 0.6146, "step": 2795 }, { "epoch": 2.1279268989149056, "grad_norm": 31.765361785888672, "learning_rate": 1.9942207986665084e-06, "loss": 0.5119, "step": 2796 }, { "epoch": 2.128688368551304, "grad_norm": 46.64960861206055, "learning_rate": 1.99099651277716e-06, "loss": 0.9026, "step": 2797 }, { "epoch": 2.129449838187702, "grad_norm": 44.1891975402832, "learning_rate": 1.987774187432517e-06, "loss": 0.8484, "step": 2798 }, { "epoch": 2.1302113078241005, "grad_norm": 34.10715866088867, "learning_rate": 1.984553824732113e-06, "loss": 0.7521, "step": 2799 }, { "epoch": 2.130972777460499, "grad_norm": 34.514915466308594, "learning_rate": 1.98133542677421e-06, "loss": 0.6402, "step": 2800 }, { "epoch": 2.131734247096897, "grad_norm": 27.983064651489258, "learning_rate": 1.9781189956557845e-06, "loss": 0.5518, "step": 2801 }, { "epoch": 2.1324957167332954, "grad_norm": 30.127460479736328, "learning_rate": 1.974904533472532e-06, "loss": 0.6462, "step": 2802 }, { "epoch": 2.1332571863696934, "grad_norm": 31.548460006713867, "learning_rate": 1.9716920423188662e-06, "loss": 0.6718, "step": 2803 }, { "epoch": 2.134018656006092, "grad_norm": 42.44682312011719, "learning_rate": 1.9684815242879156e-06, "loss": 0.7772, "step": 2804 }, { "epoch": 2.13478012564249, "grad_norm": 31.550262451171875, "learning_rate": 1.965272981471524e-06, "loss": 0.6608, "step": 2805 }, { "epoch": 2.1355415952788883, "grad_norm": 32.023109436035156, "learning_rate": 1.962066415960247e-06, "loss": 1.172, "step": 2806 }, { "epoch": 2.1363030649152863, "grad_norm": 35.00474548339844, "learning_rate": 1.958861829843354e-06, "loss": 0.5167, "step": 2807 }, { "epoch": 2.1370645345516848, "grad_norm": 29.72364044189453, "learning_rate": 1.9556592252088215e-06, "loss": 0.6874, "step": 2808 }, { "epoch": 2.137826004188083, "grad_norm": 26.90749740600586, "learning_rate": 1.9524586041433393e-06, "loss": 0.5334, "step": 2809 }, { "epoch": 2.1385874738244812, "grad_norm": 27.878686904907227, "learning_rate": 1.9492599687323012e-06, "loss": 0.6203, "step": 2810 }, { "epoch": 2.1393489434608797, "grad_norm": 37.585899353027344, "learning_rate": 1.9460633210598083e-06, "loss": 0.7551, "step": 2811 }, { "epoch": 2.1401104130972777, "grad_norm": 25.899089813232422, "learning_rate": 1.9428686632086674e-06, "loss": 0.602, "step": 2812 }, { "epoch": 2.140871882733676, "grad_norm": 25.4023380279541, "learning_rate": 1.9396759972603867e-06, "loss": 0.5237, "step": 2813 }, { "epoch": 2.141633352370074, "grad_norm": 34.30722427368164, "learning_rate": 1.936485325295179e-06, "loss": 0.8066, "step": 2814 }, { "epoch": 2.1423948220064726, "grad_norm": 31.772428512573242, "learning_rate": 1.933296649391957e-06, "loss": 0.5941, "step": 2815 }, { "epoch": 2.1431562916428706, "grad_norm": 28.733829498291016, "learning_rate": 1.9301099716283293e-06, "loss": 0.5367, "step": 2816 }, { "epoch": 2.143917761279269, "grad_norm": 34.2606086730957, "learning_rate": 1.9269252940806115e-06, "loss": 0.7358, "step": 2817 }, { "epoch": 2.144679230915667, "grad_norm": 31.52211570739746, "learning_rate": 1.9237426188238074e-06, "loss": 0.7809, "step": 2818 }, { "epoch": 2.1454407005520655, "grad_norm": 40.96409606933594, "learning_rate": 1.920561947931621e-06, "loss": 0.8137, "step": 2819 }, { "epoch": 2.146202170188464, "grad_norm": 28.828495025634766, "learning_rate": 1.917383283476444e-06, "loss": 0.5776, "step": 2820 }, { "epoch": 2.146963639824862, "grad_norm": 32.983665466308594, "learning_rate": 1.9142066275293696e-06, "loss": 0.8051, "step": 2821 }, { "epoch": 2.1477251094612604, "grad_norm": 35.42782211303711, "learning_rate": 1.911031982160177e-06, "loss": 0.6263, "step": 2822 }, { "epoch": 2.1484865790976584, "grad_norm": 27.241426467895508, "learning_rate": 1.907859349437336e-06, "loss": 0.657, "step": 2823 }, { "epoch": 2.149248048734057, "grad_norm": 28.718557357788086, "learning_rate": 1.904688731428006e-06, "loss": 0.6283, "step": 2824 }, { "epoch": 2.150009518370455, "grad_norm": 33.689292907714844, "learning_rate": 1.9015201301980313e-06, "loss": 0.6987, "step": 2825 }, { "epoch": 2.1507709880068533, "grad_norm": 32.93886947631836, "learning_rate": 1.8983535478119487e-06, "loss": 0.7101, "step": 2826 }, { "epoch": 2.1515324576432513, "grad_norm": 27.841161727905273, "learning_rate": 1.8951889863329703e-06, "loss": 0.6727, "step": 2827 }, { "epoch": 2.1522939272796497, "grad_norm": 26.885433197021484, "learning_rate": 1.8920264478229977e-06, "loss": 0.6195, "step": 2828 }, { "epoch": 2.1530553969160477, "grad_norm": 32.902095794677734, "learning_rate": 1.8888659343426108e-06, "loss": 0.705, "step": 2829 }, { "epoch": 2.153816866552446, "grad_norm": 29.938688278198242, "learning_rate": 1.8857074479510761e-06, "loss": 0.6485, "step": 2830 }, { "epoch": 2.1545783361888446, "grad_norm": 35.626434326171875, "learning_rate": 1.8825509907063328e-06, "loss": 0.8448, "step": 2831 }, { "epoch": 2.1553398058252426, "grad_norm": 37.05457305908203, "learning_rate": 1.8793965646650009e-06, "loss": 0.6941, "step": 2832 }, { "epoch": 2.156101275461641, "grad_norm": 36.70185852050781, "learning_rate": 1.8762441718823754e-06, "loss": 0.7935, "step": 2833 }, { "epoch": 2.156862745098039, "grad_norm": 27.31637191772461, "learning_rate": 1.8730938144124317e-06, "loss": 0.6315, "step": 2834 }, { "epoch": 2.1576242147344376, "grad_norm": 27.035017013549805, "learning_rate": 1.8699454943078104e-06, "loss": 0.5412, "step": 2835 }, { "epoch": 2.1583856843708356, "grad_norm": 30.69132423400879, "learning_rate": 1.8667992136198315e-06, "loss": 0.5028, "step": 2836 }, { "epoch": 2.159147154007234, "grad_norm": 39.04477310180664, "learning_rate": 1.8636549743984815e-06, "loss": 0.754, "step": 2837 }, { "epoch": 2.159908623643632, "grad_norm": 32.79206466674805, "learning_rate": 1.8605127786924225e-06, "loss": 0.6668, "step": 2838 }, { "epoch": 2.1606700932800305, "grad_norm": 26.097702026367188, "learning_rate": 1.8573726285489813e-06, "loss": 0.5118, "step": 2839 }, { "epoch": 2.161431562916429, "grad_norm": 33.49382781982422, "learning_rate": 1.8542345260141514e-06, "loss": 0.6328, "step": 2840 }, { "epoch": 2.162193032552827, "grad_norm": 29.287303924560547, "learning_rate": 1.851098473132593e-06, "loss": 0.6277, "step": 2841 }, { "epoch": 2.1629545021892254, "grad_norm": 34.785675048828125, "learning_rate": 1.8479644719476324e-06, "loss": 0.8314, "step": 2842 }, { "epoch": 2.1637159718256234, "grad_norm": 34.707420349121094, "learning_rate": 1.8448325245012567e-06, "loss": 0.6377, "step": 2843 }, { "epoch": 2.164477441462022, "grad_norm": 23.32147789001465, "learning_rate": 1.8417026328341158e-06, "loss": 0.4678, "step": 2844 }, { "epoch": 2.16523891109842, "grad_norm": 35.71901321411133, "learning_rate": 1.8385747989855206e-06, "loss": 0.674, "step": 2845 }, { "epoch": 2.1660003807348183, "grad_norm": 33.228721618652344, "learning_rate": 1.8354490249934392e-06, "loss": 0.5674, "step": 2846 }, { "epoch": 2.1667618503712163, "grad_norm": 36.07897186279297, "learning_rate": 1.8323253128945023e-06, "loss": 0.6829, "step": 2847 }, { "epoch": 2.1675233200076147, "grad_norm": 28.633350372314453, "learning_rate": 1.829203664723993e-06, "loss": 0.5452, "step": 2848 }, { "epoch": 2.168284789644013, "grad_norm": 27.89918327331543, "learning_rate": 1.8260840825158505e-06, "loss": 0.6613, "step": 2849 }, { "epoch": 2.169046259280411, "grad_norm": 35.77248764038086, "learning_rate": 1.8229665683026675e-06, "loss": 0.7206, "step": 2850 }, { "epoch": 2.1698077289168096, "grad_norm": 34.655784606933594, "learning_rate": 1.8198511241156902e-06, "loss": 0.9018, "step": 2851 }, { "epoch": 2.1705691985532076, "grad_norm": 34.06140899658203, "learning_rate": 1.8167377519848161e-06, "loss": 0.5756, "step": 2852 }, { "epoch": 2.171330668189606, "grad_norm": 33.97480010986328, "learning_rate": 1.8136264539385924e-06, "loss": 0.7065, "step": 2853 }, { "epoch": 2.172092137826004, "grad_norm": 27.486658096313477, "learning_rate": 1.8105172320042124e-06, "loss": 0.6128, "step": 2854 }, { "epoch": 2.1728536074624025, "grad_norm": 35.0899772644043, "learning_rate": 1.8074100882075229e-06, "loss": 0.7385, "step": 2855 }, { "epoch": 2.1736150770988005, "grad_norm": 32.74992370605469, "learning_rate": 1.8043050245730115e-06, "loss": 0.6897, "step": 2856 }, { "epoch": 2.174376546735199, "grad_norm": 29.50080680847168, "learning_rate": 1.801202043123811e-06, "loss": 0.6658, "step": 2857 }, { "epoch": 2.175138016371597, "grad_norm": 29.388471603393555, "learning_rate": 1.7981011458816988e-06, "loss": 0.6722, "step": 2858 }, { "epoch": 2.1758994860079954, "grad_norm": 40.552433013916016, "learning_rate": 1.7950023348670936e-06, "loss": 0.8034, "step": 2859 }, { "epoch": 2.176660955644394, "grad_norm": 29.826169967651367, "learning_rate": 1.7919056120990558e-06, "loss": 0.7444, "step": 2860 }, { "epoch": 2.177422425280792, "grad_norm": 26.280885696411133, "learning_rate": 1.7888109795952836e-06, "loss": 0.6908, "step": 2861 }, { "epoch": 2.1781838949171903, "grad_norm": 33.81587600708008, "learning_rate": 1.7857184393721145e-06, "loss": 0.8399, "step": 2862 }, { "epoch": 2.1789453645535883, "grad_norm": 35.26399612426758, "learning_rate": 1.78262799344452e-06, "loss": 0.81, "step": 2863 }, { "epoch": 2.179706834189987, "grad_norm": 36.797447204589844, "learning_rate": 1.7795396438261138e-06, "loss": 0.6825, "step": 2864 }, { "epoch": 2.180468303826385, "grad_norm": 31.9467716217041, "learning_rate": 1.776453392529139e-06, "loss": 0.7129, "step": 2865 }, { "epoch": 2.1812297734627832, "grad_norm": 37.27809524536133, "learning_rate": 1.7733692415644676e-06, "loss": 0.5892, "step": 2866 }, { "epoch": 2.1819912430991812, "grad_norm": 45.14113235473633, "learning_rate": 1.7702871929416088e-06, "loss": 0.6758, "step": 2867 }, { "epoch": 2.1827527127355797, "grad_norm": 28.038671493530273, "learning_rate": 1.7672072486687031e-06, "loss": 0.5398, "step": 2868 }, { "epoch": 2.1835141823719777, "grad_norm": 33.22085189819336, "learning_rate": 1.7641294107525152e-06, "loss": 0.5339, "step": 2869 }, { "epoch": 2.184275652008376, "grad_norm": 44.120033264160156, "learning_rate": 1.7610536811984407e-06, "loss": 0.8732, "step": 2870 }, { "epoch": 2.1850371216447746, "grad_norm": 30.33187484741211, "learning_rate": 1.7579800620104982e-06, "loss": 0.7394, "step": 2871 }, { "epoch": 2.1857985912811726, "grad_norm": 30.51215171813965, "learning_rate": 1.7549085551913358e-06, "loss": 0.6082, "step": 2872 }, { "epoch": 2.186560060917571, "grad_norm": 30.488056182861328, "learning_rate": 1.7518391627422243e-06, "loss": 0.6556, "step": 2873 }, { "epoch": 2.187321530553969, "grad_norm": 31.807451248168945, "learning_rate": 1.748771886663051e-06, "loss": 0.7602, "step": 2874 }, { "epoch": 2.1880830001903675, "grad_norm": 28.23213005065918, "learning_rate": 1.7457067289523294e-06, "loss": 0.7039, "step": 2875 }, { "epoch": 2.1888444698267655, "grad_norm": 30.43992042541504, "learning_rate": 1.7426436916071943e-06, "loss": 0.6836, "step": 2876 }, { "epoch": 2.189605939463164, "grad_norm": 31.241214752197266, "learning_rate": 1.7395827766233964e-06, "loss": 0.6717, "step": 2877 }, { "epoch": 2.190367409099562, "grad_norm": 33.621063232421875, "learning_rate": 1.7365239859953026e-06, "loss": 0.6884, "step": 2878 }, { "epoch": 2.1911288787359604, "grad_norm": 29.696969985961914, "learning_rate": 1.7334673217158976e-06, "loss": 0.5811, "step": 2879 }, { "epoch": 2.191890348372359, "grad_norm": 38.869476318359375, "learning_rate": 1.7304127857767778e-06, "loss": 0.8582, "step": 2880 }, { "epoch": 2.192651818008757, "grad_norm": 28.328012466430664, "learning_rate": 1.7273603801681604e-06, "loss": 0.5836, "step": 2881 }, { "epoch": 2.1934132876451553, "grad_norm": 34.223045349121094, "learning_rate": 1.7243101068788642e-06, "loss": 0.7213, "step": 2882 }, { "epoch": 2.1941747572815533, "grad_norm": 32.9385871887207, "learning_rate": 1.7212619678963255e-06, "loss": 0.6216, "step": 2883 }, { "epoch": 2.1949362269179518, "grad_norm": 34.677040100097656, "learning_rate": 1.7182159652065855e-06, "loss": 0.6326, "step": 2884 }, { "epoch": 2.1956976965543498, "grad_norm": 33.123634338378906, "learning_rate": 1.7151721007943e-06, "loss": 0.6774, "step": 2885 }, { "epoch": 2.196459166190748, "grad_norm": 28.249923706054688, "learning_rate": 1.7121303766427266e-06, "loss": 0.6758, "step": 2886 }, { "epoch": 2.197220635827146, "grad_norm": 30.19939613342285, "learning_rate": 1.7090907947337287e-06, "loss": 0.6338, "step": 2887 }, { "epoch": 2.1979821054635447, "grad_norm": 29.172266006469727, "learning_rate": 1.7060533570477728e-06, "loss": 0.5362, "step": 2888 }, { "epoch": 2.198743575099943, "grad_norm": 40.944122314453125, "learning_rate": 1.7030180655639361e-06, "loss": 0.8336, "step": 2889 }, { "epoch": 2.199505044736341, "grad_norm": 28.077306747436523, "learning_rate": 1.6999849222598851e-06, "loss": 0.7072, "step": 2890 }, { "epoch": 2.2002665143727396, "grad_norm": 27.57504653930664, "learning_rate": 1.6969539291118963e-06, "loss": 0.5063, "step": 2891 }, { "epoch": 2.2010279840091376, "grad_norm": 35.02766418457031, "learning_rate": 1.6939250880948393e-06, "loss": 0.5823, "step": 2892 }, { "epoch": 2.201789453645536, "grad_norm": 38.76290512084961, "learning_rate": 1.6908984011821883e-06, "loss": 0.7888, "step": 2893 }, { "epoch": 2.202550923281934, "grad_norm": 32.44579315185547, "learning_rate": 1.6878738703460069e-06, "loss": 0.6725, "step": 2894 }, { "epoch": 2.2033123929183325, "grad_norm": 40.927547454833984, "learning_rate": 1.6848514975569575e-06, "loss": 0.8525, "step": 2895 }, { "epoch": 2.2040738625547305, "grad_norm": 31.171796798706055, "learning_rate": 1.681831284784296e-06, "loss": 0.631, "step": 2896 }, { "epoch": 2.204835332191129, "grad_norm": 28.894502639770508, "learning_rate": 1.6788132339958701e-06, "loss": 0.6541, "step": 2897 }, { "epoch": 2.205596801827527, "grad_norm": 37.0275993347168, "learning_rate": 1.6757973471581196e-06, "loss": 0.6143, "step": 2898 }, { "epoch": 2.2063582714639254, "grad_norm": 51.29191207885742, "learning_rate": 1.6727836262360742e-06, "loss": 0.8999, "step": 2899 }, { "epoch": 2.207119741100324, "grad_norm": 34.941864013671875, "learning_rate": 1.669772073193352e-06, "loss": 0.7705, "step": 2900 }, { "epoch": 2.207881210736722, "grad_norm": 36.182037353515625, "learning_rate": 1.6667626899921574e-06, "loss": 0.7683, "step": 2901 }, { "epoch": 2.2086426803731203, "grad_norm": 32.4496955871582, "learning_rate": 1.6637554785932857e-06, "loss": 0.8367, "step": 2902 }, { "epoch": 2.2094041500095183, "grad_norm": 37.02867889404297, "learning_rate": 1.6607504409561126e-06, "loss": 0.711, "step": 2903 }, { "epoch": 2.2101656196459167, "grad_norm": 39.95683288574219, "learning_rate": 1.6577475790385984e-06, "loss": 0.6376, "step": 2904 }, { "epoch": 2.2109270892823147, "grad_norm": 34.53549575805664, "learning_rate": 1.6547468947972866e-06, "loss": 0.6767, "step": 2905 }, { "epoch": 2.211688558918713, "grad_norm": 25.173511505126953, "learning_rate": 1.6517483901873016e-06, "loss": 0.5058, "step": 2906 }, { "epoch": 2.212450028555111, "grad_norm": 36.99832534790039, "learning_rate": 1.6487520671623469e-06, "loss": 0.7521, "step": 2907 }, { "epoch": 2.2132114981915096, "grad_norm": 31.540210723876953, "learning_rate": 1.6457579276747049e-06, "loss": 0.6139, "step": 2908 }, { "epoch": 2.2139729678279076, "grad_norm": 41.95873260498047, "learning_rate": 1.6427659736752349e-06, "loss": 0.7035, "step": 2909 }, { "epoch": 2.214734437464306, "grad_norm": 34.704978942871094, "learning_rate": 1.6397762071133755e-06, "loss": 0.6035, "step": 2910 }, { "epoch": 2.2154959071007045, "grad_norm": 31.481212615966797, "learning_rate": 1.6367886299371355e-06, "loss": 0.642, "step": 2911 }, { "epoch": 2.2162573767371025, "grad_norm": 39.797122955322266, "learning_rate": 1.6338032440931017e-06, "loss": 0.7714, "step": 2912 }, { "epoch": 2.217018846373501, "grad_norm": 40.132843017578125, "learning_rate": 1.6308200515264243e-06, "loss": 0.6755, "step": 2913 }, { "epoch": 2.217780316009899, "grad_norm": 38.351619720458984, "learning_rate": 1.6278390541808364e-06, "loss": 0.8081, "step": 2914 }, { "epoch": 2.2185417856462974, "grad_norm": 37.40230941772461, "learning_rate": 1.6248602539986337e-06, "loss": 0.7814, "step": 2915 }, { "epoch": 2.2193032552826955, "grad_norm": 25.124324798583984, "learning_rate": 1.6218836529206804e-06, "loss": 0.6325, "step": 2916 }, { "epoch": 2.220064724919094, "grad_norm": 36.72060012817383, "learning_rate": 1.6189092528864104e-06, "loss": 0.5967, "step": 2917 }, { "epoch": 2.220826194555492, "grad_norm": 39.21923828125, "learning_rate": 1.61593705583382e-06, "loss": 0.7511, "step": 2918 }, { "epoch": 2.2215876641918904, "grad_norm": 35.650962829589844, "learning_rate": 1.612967063699476e-06, "loss": 0.7085, "step": 2919 }, { "epoch": 2.222349133828289, "grad_norm": 37.90719985961914, "learning_rate": 1.6099992784185037e-06, "loss": 0.7353, "step": 2920 }, { "epoch": 2.223110603464687, "grad_norm": 27.398759841918945, "learning_rate": 1.6070337019245896e-06, "loss": 0.605, "step": 2921 }, { "epoch": 2.2238720731010853, "grad_norm": 38.77033996582031, "learning_rate": 1.6040703361499833e-06, "loss": 0.7063, "step": 2922 }, { "epoch": 2.2246335427374833, "grad_norm": 40.20273971557617, "learning_rate": 1.601109183025496e-06, "loss": 0.8809, "step": 2923 }, { "epoch": 2.2253950123738817, "grad_norm": 37.87397384643555, "learning_rate": 1.5981502444804942e-06, "loss": 0.6478, "step": 2924 }, { "epoch": 2.2261564820102797, "grad_norm": 26.36509132385254, "learning_rate": 1.595193522442902e-06, "loss": 0.547, "step": 2925 }, { "epoch": 2.226917951646678, "grad_norm": 34.07435607910156, "learning_rate": 1.5922390188391984e-06, "loss": 0.5863, "step": 2926 }, { "epoch": 2.227679421283076, "grad_norm": 36.228515625, "learning_rate": 1.5892867355944209e-06, "loss": 0.7364, "step": 2927 }, { "epoch": 2.2284408909194746, "grad_norm": 30.74782943725586, "learning_rate": 1.5863366746321578e-06, "loss": 0.5058, "step": 2928 }, { "epoch": 2.229202360555873, "grad_norm": 29.14112091064453, "learning_rate": 1.5833888378745465e-06, "loss": 0.5699, "step": 2929 }, { "epoch": 2.229963830192271, "grad_norm": 27.05422019958496, "learning_rate": 1.580443227242277e-06, "loss": 0.5228, "step": 2930 }, { "epoch": 2.2307252998286695, "grad_norm": 42.71877670288086, "learning_rate": 1.5774998446545936e-06, "loss": 0.7706, "step": 2931 }, { "epoch": 2.2314867694650675, "grad_norm": 30.529537200927734, "learning_rate": 1.5745586920292838e-06, "loss": 0.7479, "step": 2932 }, { "epoch": 2.232248239101466, "grad_norm": 37.62065505981445, "learning_rate": 1.5716197712826826e-06, "loss": 0.7888, "step": 2933 }, { "epoch": 2.233009708737864, "grad_norm": 43.42455291748047, "learning_rate": 1.5686830843296724e-06, "loss": 0.845, "step": 2934 }, { "epoch": 2.2337711783742624, "grad_norm": 39.82248306274414, "learning_rate": 1.5657486330836786e-06, "loss": 0.8174, "step": 2935 }, { "epoch": 2.2345326480106604, "grad_norm": 28.97212028503418, "learning_rate": 1.5628164194566713e-06, "loss": 0.7269, "step": 2936 }, { "epoch": 2.235294117647059, "grad_norm": 28.25875473022461, "learning_rate": 1.5598864453591622e-06, "loss": 0.6219, "step": 2937 }, { "epoch": 2.236055587283457, "grad_norm": 30.240699768066406, "learning_rate": 1.5569587127002034e-06, "loss": 0.57, "step": 2938 }, { "epoch": 2.2368170569198553, "grad_norm": 32.85457229614258, "learning_rate": 1.5540332233873846e-06, "loss": 0.7483, "step": 2939 }, { "epoch": 2.2375785265562538, "grad_norm": 31.80710220336914, "learning_rate": 1.5511099793268403e-06, "loss": 0.681, "step": 2940 }, { "epoch": 2.238339996192652, "grad_norm": 27.60805892944336, "learning_rate": 1.5481889824232354e-06, "loss": 0.7448, "step": 2941 }, { "epoch": 2.2391014658290502, "grad_norm": 25.262500762939453, "learning_rate": 1.5452702345797738e-06, "loss": 0.5032, "step": 2942 }, { "epoch": 2.2398629354654482, "grad_norm": 38.41075134277344, "learning_rate": 1.5423537376981924e-06, "loss": 0.8247, "step": 2943 }, { "epoch": 2.2406244051018467, "grad_norm": 24.413249969482422, "learning_rate": 1.5394394936787631e-06, "loss": 0.5049, "step": 2944 }, { "epoch": 2.2413858747382447, "grad_norm": 29.19280242919922, "learning_rate": 1.5365275044202893e-06, "loss": 0.5167, "step": 2945 }, { "epoch": 2.242147344374643, "grad_norm": 33.994380950927734, "learning_rate": 1.5336177718201045e-06, "loss": 0.7761, "step": 2946 }, { "epoch": 2.242908814011041, "grad_norm": 28.2944393157959, "learning_rate": 1.5307102977740712e-06, "loss": 0.6187, "step": 2947 }, { "epoch": 2.2436702836474396, "grad_norm": 35.285030364990234, "learning_rate": 1.5278050841765852e-06, "loss": 0.7126, "step": 2948 }, { "epoch": 2.2444317532838376, "grad_norm": 31.43611717224121, "learning_rate": 1.5249021329205638e-06, "loss": 0.5764, "step": 2949 }, { "epoch": 2.245193222920236, "grad_norm": 32.06356430053711, "learning_rate": 1.5220014458974524e-06, "loss": 0.6263, "step": 2950 }, { "epoch": 2.2459546925566345, "grad_norm": 32.95655822753906, "learning_rate": 1.5191030249972212e-06, "loss": 0.6477, "step": 2951 }, { "epoch": 2.2467161621930325, "grad_norm": 31.19467544555664, "learning_rate": 1.516206872108364e-06, "loss": 0.7839, "step": 2952 }, { "epoch": 2.247477631829431, "grad_norm": 29.707521438598633, "learning_rate": 1.5133129891178966e-06, "loss": 0.709, "step": 2953 }, { "epoch": 2.248239101465829, "grad_norm": 34.66994094848633, "learning_rate": 1.5104213779113553e-06, "loss": 0.7048, "step": 2954 }, { "epoch": 2.2490005711022274, "grad_norm": 31.111833572387695, "learning_rate": 1.5075320403727978e-06, "loss": 0.732, "step": 2955 }, { "epoch": 2.2497620407386254, "grad_norm": 32.51296615600586, "learning_rate": 1.5046449783847965e-06, "loss": 0.7312, "step": 2956 }, { "epoch": 2.250523510375024, "grad_norm": 36.1729850769043, "learning_rate": 1.501760193828448e-06, "loss": 0.8427, "step": 2957 }, { "epoch": 2.251284980011422, "grad_norm": 35.177642822265625, "learning_rate": 1.498877688583359e-06, "loss": 0.6866, "step": 2958 }, { "epoch": 2.2520464496478203, "grad_norm": 37.65130615234375, "learning_rate": 1.4959974645276548e-06, "loss": 0.5776, "step": 2959 }, { "epoch": 2.2528079192842183, "grad_norm": 33.416542053222656, "learning_rate": 1.4931195235379681e-06, "loss": 0.7569, "step": 2960 }, { "epoch": 2.2535693889206168, "grad_norm": 42.56591796875, "learning_rate": 1.4902438674894533e-06, "loss": 0.9529, "step": 2961 }, { "epoch": 2.254330858557015, "grad_norm": 33.28287124633789, "learning_rate": 1.487370498255769e-06, "loss": 0.7523, "step": 2962 }, { "epoch": 2.255092328193413, "grad_norm": 32.36593246459961, "learning_rate": 1.4844994177090871e-06, "loss": 0.794, "step": 2963 }, { "epoch": 2.2558537978298117, "grad_norm": 31.53890037536621, "learning_rate": 1.4816306277200853e-06, "loss": 0.6513, "step": 2964 }, { "epoch": 2.2566152674662097, "grad_norm": 35.16249084472656, "learning_rate": 1.4787641301579541e-06, "loss": 0.6164, "step": 2965 }, { "epoch": 2.257376737102608, "grad_norm": 27.199729919433594, "learning_rate": 1.475899926890385e-06, "loss": 0.5705, "step": 2966 }, { "epoch": 2.258138206739006, "grad_norm": 20.075571060180664, "learning_rate": 1.4730380197835792e-06, "loss": 0.4275, "step": 2967 }, { "epoch": 2.2588996763754046, "grad_norm": 43.94169998168945, "learning_rate": 1.4701784107022343e-06, "loss": 0.8228, "step": 2968 }, { "epoch": 2.259661146011803, "grad_norm": 32.405704498291016, "learning_rate": 1.4673211015095607e-06, "loss": 0.5857, "step": 2969 }, { "epoch": 2.260422615648201, "grad_norm": 30.990280151367188, "learning_rate": 1.4644660940672628e-06, "loss": 0.555, "step": 2970 }, { "epoch": 2.2611840852845995, "grad_norm": 28.03704261779785, "learning_rate": 1.4616133902355484e-06, "loss": 0.6027, "step": 2971 }, { "epoch": 2.2619455549209975, "grad_norm": 39.21113967895508, "learning_rate": 1.458762991873124e-06, "loss": 0.6775, "step": 2972 }, { "epoch": 2.262707024557396, "grad_norm": 35.101924896240234, "learning_rate": 1.4559149008371908e-06, "loss": 0.8569, "step": 2973 }, { "epoch": 2.263468494193794, "grad_norm": 35.61239242553711, "learning_rate": 1.4530691189834555e-06, "loss": 0.5018, "step": 2974 }, { "epoch": 2.2642299638301924, "grad_norm": 35.984901428222656, "learning_rate": 1.4502256481661086e-06, "loss": 0.6091, "step": 2975 }, { "epoch": 2.2649914334665904, "grad_norm": 30.734619140625, "learning_rate": 1.4473844902378426e-06, "loss": 0.671, "step": 2976 }, { "epoch": 2.265752903102989, "grad_norm": 33.42219924926758, "learning_rate": 1.4445456470498392e-06, "loss": 0.7016, "step": 2977 }, { "epoch": 2.266514372739387, "grad_norm": 45.224937438964844, "learning_rate": 1.4417091204517764e-06, "loss": 0.7906, "step": 2978 }, { "epoch": 2.2672758423757853, "grad_norm": 37.9051628112793, "learning_rate": 1.4388749122918189e-06, "loss": 0.613, "step": 2979 }, { "epoch": 2.2680373120121837, "grad_norm": 29.646198272705078, "learning_rate": 1.4360430244166207e-06, "loss": 0.7766, "step": 2980 }, { "epoch": 2.2687987816485817, "grad_norm": 35.6751823425293, "learning_rate": 1.4332134586713253e-06, "loss": 0.783, "step": 2981 }, { "epoch": 2.26956025128498, "grad_norm": 36.23665237426758, "learning_rate": 1.4303862168995664e-06, "loss": 0.8805, "step": 2982 }, { "epoch": 2.270321720921378, "grad_norm": 39.54021072387695, "learning_rate": 1.4275613009434553e-06, "loss": 0.7861, "step": 2983 }, { "epoch": 2.2710831905577766, "grad_norm": 36.523807525634766, "learning_rate": 1.4247387126435957e-06, "loss": 0.6644, "step": 2984 }, { "epoch": 2.2718446601941746, "grad_norm": 33.768646240234375, "learning_rate": 1.421918453839069e-06, "loss": 0.6398, "step": 2985 }, { "epoch": 2.272606129830573, "grad_norm": 31.91346549987793, "learning_rate": 1.4191005263674445e-06, "loss": 0.5818, "step": 2986 }, { "epoch": 2.273367599466971, "grad_norm": 33.07283020019531, "learning_rate": 1.4162849320647693e-06, "loss": 0.6815, "step": 2987 }, { "epoch": 2.2741290691033695, "grad_norm": 43.88042449951172, "learning_rate": 1.413471672765569e-06, "loss": 0.6706, "step": 2988 }, { "epoch": 2.2748905387397675, "grad_norm": 34.32038879394531, "learning_rate": 1.4106607503028503e-06, "loss": 0.5639, "step": 2989 }, { "epoch": 2.275652008376166, "grad_norm": 31.32863426208496, "learning_rate": 1.4078521665080957e-06, "loss": 0.7314, "step": 2990 }, { "epoch": 2.2764134780125644, "grad_norm": 32.91825485229492, "learning_rate": 1.4050459232112652e-06, "loss": 0.6435, "step": 2991 }, { "epoch": 2.2771749476489624, "grad_norm": 33.36286163330078, "learning_rate": 1.4022420222407923e-06, "loss": 0.6525, "step": 2992 }, { "epoch": 2.277936417285361, "grad_norm": 31.652034759521484, "learning_rate": 1.3994404654235866e-06, "loss": 0.7814, "step": 2993 }, { "epoch": 2.278697886921759, "grad_norm": 27.62053871154785, "learning_rate": 1.396641254585026e-06, "loss": 0.6213, "step": 2994 }, { "epoch": 2.2794593565581573, "grad_norm": 38.289493560791016, "learning_rate": 1.3938443915489664e-06, "loss": 0.788, "step": 2995 }, { "epoch": 2.2802208261945554, "grad_norm": 30.05084800720215, "learning_rate": 1.391049878137729e-06, "loss": 0.7371, "step": 2996 }, { "epoch": 2.280982295830954, "grad_norm": 33.213401794433594, "learning_rate": 1.3882577161721055e-06, "loss": 0.7603, "step": 2997 }, { "epoch": 2.281743765467352, "grad_norm": 27.739910125732422, "learning_rate": 1.3854679074713557e-06, "loss": 0.6267, "step": 2998 }, { "epoch": 2.2825052351037503, "grad_norm": 33.543190002441406, "learning_rate": 1.3826804538532051e-06, "loss": 0.717, "step": 2999 }, { "epoch": 2.2832667047401483, "grad_norm": 31.803388595581055, "learning_rate": 1.3798953571338457e-06, "loss": 0.6004, "step": 3000 }, { "epoch": 2.2840281743765467, "grad_norm": 37.393341064453125, "learning_rate": 1.3771126191279344e-06, "loss": 0.8278, "step": 3001 }, { "epoch": 2.284789644012945, "grad_norm": 29.509075164794922, "learning_rate": 1.3743322416485882e-06, "loss": 0.6509, "step": 3002 }, { "epoch": 2.285551113649343, "grad_norm": 26.711210250854492, "learning_rate": 1.3715542265073917e-06, "loss": 0.614, "step": 3003 }, { "epoch": 2.2863125832857416, "grad_norm": 33.4382438659668, "learning_rate": 1.3687785755143851e-06, "loss": 0.6572, "step": 3004 }, { "epoch": 2.2870740529221396, "grad_norm": 25.8206844329834, "learning_rate": 1.3660052904780707e-06, "loss": 0.5573, "step": 3005 }, { "epoch": 2.287835522558538, "grad_norm": 34.00285720825195, "learning_rate": 1.3632343732054076e-06, "loss": 0.6497, "step": 3006 }, { "epoch": 2.288596992194936, "grad_norm": 39.669883728027344, "learning_rate": 1.3604658255018145e-06, "loss": 0.7681, "step": 3007 }, { "epoch": 2.2893584618313345, "grad_norm": 28.40967559814453, "learning_rate": 1.3576996491711642e-06, "loss": 0.6161, "step": 3008 }, { "epoch": 2.290119931467733, "grad_norm": 31.365480422973633, "learning_rate": 1.3549358460157846e-06, "loss": 0.7086, "step": 3009 }, { "epoch": 2.290881401104131, "grad_norm": 36.632423400878906, "learning_rate": 1.3521744178364588e-06, "loss": 0.6374, "step": 3010 }, { "epoch": 2.2916428707405294, "grad_norm": 31.29145050048828, "learning_rate": 1.3494153664324188e-06, "loss": 0.6877, "step": 3011 }, { "epoch": 2.2924043403769274, "grad_norm": 41.93159484863281, "learning_rate": 1.3466586936013548e-06, "loss": 0.7387, "step": 3012 }, { "epoch": 2.293165810013326, "grad_norm": 38.682979583740234, "learning_rate": 1.3439044011394025e-06, "loss": 0.7861, "step": 3013 }, { "epoch": 2.293927279649724, "grad_norm": 34.934730529785156, "learning_rate": 1.341152490841144e-06, "loss": 0.6122, "step": 3014 }, { "epoch": 2.2946887492861223, "grad_norm": 35.781044006347656, "learning_rate": 1.3384029644996132e-06, "loss": 0.7438, "step": 3015 }, { "epoch": 2.2954502189225203, "grad_norm": 31.7769718170166, "learning_rate": 1.3356558239062921e-06, "loss": 0.6496, "step": 3016 }, { "epoch": 2.2962116885589188, "grad_norm": 34.74313735961914, "learning_rate": 1.3329110708511062e-06, "loss": 0.7071, "step": 3017 }, { "epoch": 2.2969731581953168, "grad_norm": 33.767662048339844, "learning_rate": 1.3301687071224234e-06, "loss": 0.6729, "step": 3018 }, { "epoch": 2.2977346278317152, "grad_norm": 36.72161102294922, "learning_rate": 1.3274287345070564e-06, "loss": 0.7796, "step": 3019 }, { "epoch": 2.2984960974681137, "grad_norm": 33.742591857910156, "learning_rate": 1.324691154790263e-06, "loss": 0.8325, "step": 3020 }, { "epoch": 2.2992575671045117, "grad_norm": 31.467844009399414, "learning_rate": 1.3219559697557394e-06, "loss": 0.5007, "step": 3021 }, { "epoch": 2.30001903674091, "grad_norm": 37.08271408081055, "learning_rate": 1.3192231811856183e-06, "loss": 0.8572, "step": 3022 }, { "epoch": 2.300780506377308, "grad_norm": 37.59956741333008, "learning_rate": 1.3164927908604736e-06, "loss": 0.8999, "step": 3023 }, { "epoch": 2.3015419760137066, "grad_norm": 30.919456481933594, "learning_rate": 1.31376480055932e-06, "loss": 0.6417, "step": 3024 }, { "epoch": 2.3023034456501046, "grad_norm": 40.8868408203125, "learning_rate": 1.3110392120596044e-06, "loss": 0.6361, "step": 3025 }, { "epoch": 2.303064915286503, "grad_norm": 39.19066619873047, "learning_rate": 1.3083160271372092e-06, "loss": 0.766, "step": 3026 }, { "epoch": 2.303826384922901, "grad_norm": 30.518577575683594, "learning_rate": 1.3055952475664512e-06, "loss": 0.7092, "step": 3027 }, { "epoch": 2.3045878545592995, "grad_norm": 34.66339874267578, "learning_rate": 1.30287687512008e-06, "loss": 0.7558, "step": 3028 }, { "epoch": 2.3053493241956975, "grad_norm": 35.416465759277344, "learning_rate": 1.3001609115692792e-06, "loss": 0.6327, "step": 3029 }, { "epoch": 2.306110793832096, "grad_norm": 44.22541046142578, "learning_rate": 1.2974473586836578e-06, "loss": 0.8104, "step": 3030 }, { "epoch": 2.3068722634684944, "grad_norm": 33.7258186340332, "learning_rate": 1.2947362182312577e-06, "loss": 0.745, "step": 3031 }, { "epoch": 2.3076337331048924, "grad_norm": 33.33492660522461, "learning_rate": 1.2920274919785468e-06, "loss": 0.7814, "step": 3032 }, { "epoch": 2.308395202741291, "grad_norm": 31.209909439086914, "learning_rate": 1.2893211816904243e-06, "loss": 0.5225, "step": 3033 }, { "epoch": 2.309156672377689, "grad_norm": 34.1356201171875, "learning_rate": 1.2866172891302104e-06, "loss": 0.6786, "step": 3034 }, { "epoch": 2.3099181420140873, "grad_norm": 33.32741165161133, "learning_rate": 1.2839158160596526e-06, "loss": 0.7058, "step": 3035 }, { "epoch": 2.3106796116504853, "grad_norm": 32.41280746459961, "learning_rate": 1.2812167642389206e-06, "loss": 0.6346, "step": 3036 }, { "epoch": 2.3114410812868837, "grad_norm": 31.662927627563477, "learning_rate": 1.2785201354266075e-06, "loss": 0.6849, "step": 3037 }, { "epoch": 2.3122025509232818, "grad_norm": 31.09730339050293, "learning_rate": 1.2758259313797273e-06, "loss": 0.6696, "step": 3038 }, { "epoch": 2.31296402055968, "grad_norm": 37.102535247802734, "learning_rate": 1.273134153853714e-06, "loss": 0.519, "step": 3039 }, { "epoch": 2.313725490196078, "grad_norm": 29.679636001586914, "learning_rate": 1.2704448046024192e-06, "loss": 0.8002, "step": 3040 }, { "epoch": 2.3144869598324767, "grad_norm": 32.1679801940918, "learning_rate": 1.267757885378117e-06, "loss": 0.6983, "step": 3041 }, { "epoch": 2.315248429468875, "grad_norm": 33.56468963623047, "learning_rate": 1.265073397931494e-06, "loss": 0.691, "step": 3042 }, { "epoch": 2.316009899105273, "grad_norm": 34.68716049194336, "learning_rate": 1.2623913440116525e-06, "loss": 0.8272, "step": 3043 }, { "epoch": 2.3167713687416716, "grad_norm": 37.15320587158203, "learning_rate": 1.2597117253661106e-06, "loss": 0.761, "step": 3044 }, { "epoch": 2.3175328383780696, "grad_norm": 37.41352844238281, "learning_rate": 1.2570345437407994e-06, "loss": 0.7185, "step": 3045 }, { "epoch": 2.318294308014468, "grad_norm": 43.109832763671875, "learning_rate": 1.2543598008800612e-06, "loss": 0.926, "step": 3046 }, { "epoch": 2.319055777650866, "grad_norm": 43.485172271728516, "learning_rate": 1.2516874985266508e-06, "loss": 0.7313, "step": 3047 }, { "epoch": 2.3198172472872645, "grad_norm": 29.899946212768555, "learning_rate": 1.2490176384217323e-06, "loss": 0.5306, "step": 3048 }, { "epoch": 2.320578716923663, "grad_norm": 42.147403717041016, "learning_rate": 1.2463502223048756e-06, "loss": 0.7358, "step": 3049 }, { "epoch": 2.321340186560061, "grad_norm": 31.01053237915039, "learning_rate": 1.2436852519140652e-06, "loss": 0.6209, "step": 3050 }, { "epoch": 2.322101656196459, "grad_norm": 32.772132873535156, "learning_rate": 1.2410227289856858e-06, "loss": 0.8354, "step": 3051 }, { "epoch": 2.3228631258328574, "grad_norm": 31.236787796020508, "learning_rate": 1.2383626552545303e-06, "loss": 0.6242, "step": 3052 }, { "epoch": 2.323624595469256, "grad_norm": 31.616579055786133, "learning_rate": 1.2357050324537917e-06, "loss": 0.6734, "step": 3053 }, { "epoch": 2.324386065105654, "grad_norm": 30.00144386291504, "learning_rate": 1.2330498623150722e-06, "loss": 0.5953, "step": 3054 }, { "epoch": 2.3251475347420523, "grad_norm": 34.163089752197266, "learning_rate": 1.2303971465683734e-06, "loss": 0.5812, "step": 3055 }, { "epoch": 2.3259090043784503, "grad_norm": 32.76923370361328, "learning_rate": 1.2277468869420956e-06, "loss": 0.5924, "step": 3056 }, { "epoch": 2.3266704740148487, "grad_norm": 27.306629180908203, "learning_rate": 1.2250990851630402e-06, "loss": 0.5327, "step": 3057 }, { "epoch": 2.3274319436512467, "grad_norm": 33.91305923461914, "learning_rate": 1.2224537429564097e-06, "loss": 0.7188, "step": 3058 }, { "epoch": 2.328193413287645, "grad_norm": 34.53565216064453, "learning_rate": 1.2198108620458016e-06, "loss": 0.6866, "step": 3059 }, { "epoch": 2.3289548829240436, "grad_norm": 34.42800521850586, "learning_rate": 1.2171704441532102e-06, "loss": 0.6633, "step": 3060 }, { "epoch": 2.3297163525604416, "grad_norm": 38.31166458129883, "learning_rate": 1.2145324909990202e-06, "loss": 0.5588, "step": 3061 }, { "epoch": 2.33047782219684, "grad_norm": 36.056331634521484, "learning_rate": 1.2118970043020205e-06, "loss": 0.854, "step": 3062 }, { "epoch": 2.331239291833238, "grad_norm": 36.718994140625, "learning_rate": 1.2092639857793852e-06, "loss": 0.6632, "step": 3063 }, { "epoch": 2.3320007614696365, "grad_norm": 36.56869888305664, "learning_rate": 1.206633437146683e-06, "loss": 0.72, "step": 3064 }, { "epoch": 2.3327622311060345, "grad_norm": 35.21123504638672, "learning_rate": 1.2040053601178725e-06, "loss": 0.7125, "step": 3065 }, { "epoch": 2.333523700742433, "grad_norm": 39.656124114990234, "learning_rate": 1.2013797564053003e-06, "loss": 0.7212, "step": 3066 }, { "epoch": 2.334285170378831, "grad_norm": 31.73805809020996, "learning_rate": 1.1987566277197065e-06, "loss": 0.6441, "step": 3067 }, { "epoch": 2.3350466400152294, "grad_norm": 26.13256072998047, "learning_rate": 1.1961359757702151e-06, "loss": 0.4478, "step": 3068 }, { "epoch": 2.3358081096516274, "grad_norm": 33.509891510009766, "learning_rate": 1.1935178022643345e-06, "loss": 0.667, "step": 3069 }, { "epoch": 2.336569579288026, "grad_norm": 33.18548583984375, "learning_rate": 1.1909021089079604e-06, "loss": 0.7422, "step": 3070 }, { "epoch": 2.3373310489244243, "grad_norm": 39.25070571899414, "learning_rate": 1.1882888974053753e-06, "loss": 0.8209, "step": 3071 }, { "epoch": 2.3380925185608223, "grad_norm": 33.39265823364258, "learning_rate": 1.1856781694592395e-06, "loss": 0.6114, "step": 3072 }, { "epoch": 2.338853988197221, "grad_norm": 36.91401672363281, "learning_rate": 1.1830699267705987e-06, "loss": 0.7287, "step": 3073 }, { "epoch": 2.339615457833619, "grad_norm": 27.732175827026367, "learning_rate": 1.1804641710388764e-06, "loss": 0.5477, "step": 3074 }, { "epoch": 2.3403769274700172, "grad_norm": 36.416194915771484, "learning_rate": 1.1778609039618804e-06, "loss": 0.4931, "step": 3075 }, { "epoch": 2.3411383971064152, "grad_norm": 33.70228576660156, "learning_rate": 1.1752601272357933e-06, "loss": 0.6373, "step": 3076 }, { "epoch": 2.3418998667428137, "grad_norm": 29.470962524414062, "learning_rate": 1.1726618425551734e-06, "loss": 0.6432, "step": 3077 }, { "epoch": 2.3426613363792117, "grad_norm": 32.350101470947266, "learning_rate": 1.170066051612958e-06, "loss": 0.7338, "step": 3078 }, { "epoch": 2.34342280601561, "grad_norm": 30.706684112548828, "learning_rate": 1.1674727561004618e-06, "loss": 0.6215, "step": 3079 }, { "epoch": 2.344184275652008, "grad_norm": 33.30424118041992, "learning_rate": 1.1648819577073694e-06, "loss": 0.71, "step": 3080 }, { "epoch": 2.3449457452884066, "grad_norm": 32.32139587402344, "learning_rate": 1.1622936581217404e-06, "loss": 0.62, "step": 3081 }, { "epoch": 2.345707214924805, "grad_norm": 35.6346321105957, "learning_rate": 1.1597078590300054e-06, "loss": 0.5189, "step": 3082 }, { "epoch": 2.346468684561203, "grad_norm": 36.985870361328125, "learning_rate": 1.1571245621169668e-06, "loss": 0.7268, "step": 3083 }, { "epoch": 2.3472301541976015, "grad_norm": 33.120079040527344, "learning_rate": 1.1545437690657962e-06, "loss": 0.7446, "step": 3084 }, { "epoch": 2.3479916238339995, "grad_norm": 24.73973846435547, "learning_rate": 1.1519654815580333e-06, "loss": 0.6122, "step": 3085 }, { "epoch": 2.348753093470398, "grad_norm": 29.46769905090332, "learning_rate": 1.1493897012735861e-06, "loss": 0.6637, "step": 3086 }, { "epoch": 2.349514563106796, "grad_norm": 37.682830810546875, "learning_rate": 1.1468164298907269e-06, "loss": 0.6567, "step": 3087 }, { "epoch": 2.3502760327431944, "grad_norm": 33.2510986328125, "learning_rate": 1.144245669086098e-06, "loss": 0.6787, "step": 3088 }, { "epoch": 2.351037502379593, "grad_norm": 33.806556701660156, "learning_rate": 1.1416774205347015e-06, "loss": 0.7548, "step": 3089 }, { "epoch": 2.351798972015991, "grad_norm": 33.74296951293945, "learning_rate": 1.139111685909904e-06, "loss": 0.5452, "step": 3090 }, { "epoch": 2.352560441652389, "grad_norm": 35.76437759399414, "learning_rate": 1.1365484668834343e-06, "loss": 0.7993, "step": 3091 }, { "epoch": 2.3533219112887873, "grad_norm": 33.098663330078125, "learning_rate": 1.1339877651253823e-06, "loss": 0.5883, "step": 3092 }, { "epoch": 2.3540833809251858, "grad_norm": 29.76606559753418, "learning_rate": 1.1314295823041965e-06, "loss": 0.6763, "step": 3093 }, { "epoch": 2.3548448505615838, "grad_norm": 30.04893684387207, "learning_rate": 1.1288739200866861e-06, "loss": 0.6052, "step": 3094 }, { "epoch": 2.355606320197982, "grad_norm": 42.060752868652344, "learning_rate": 1.1263207801380155e-06, "loss": 0.896, "step": 3095 }, { "epoch": 2.35636778983438, "grad_norm": 37.12820816040039, "learning_rate": 1.1237701641217097e-06, "loss": 0.7092, "step": 3096 }, { "epoch": 2.3571292594707787, "grad_norm": 25.518049240112305, "learning_rate": 1.1212220736996455e-06, "loss": 0.4448, "step": 3097 }, { "epoch": 2.3578907291071767, "grad_norm": 28.972599029541016, "learning_rate": 1.1186765105320552e-06, "loss": 0.4674, "step": 3098 }, { "epoch": 2.358652198743575, "grad_norm": 30.620302200317383, "learning_rate": 1.1161334762775245e-06, "loss": 0.6365, "step": 3099 }, { "epoch": 2.3594136683799736, "grad_norm": 45.65397262573242, "learning_rate": 1.1135929725929912e-06, "loss": 0.8641, "step": 3100 }, { "epoch": 2.3601751380163716, "grad_norm": 32.80698013305664, "learning_rate": 1.111055001133745e-06, "loss": 0.6589, "step": 3101 }, { "epoch": 2.36093660765277, "grad_norm": 41.75936508178711, "learning_rate": 1.1085195635534257e-06, "loss": 0.6625, "step": 3102 }, { "epoch": 2.361698077289168, "grad_norm": 38.34126281738281, "learning_rate": 1.1059866615040205e-06, "loss": 0.7132, "step": 3103 }, { "epoch": 2.3624595469255665, "grad_norm": 31.895259857177734, "learning_rate": 1.1034562966358653e-06, "loss": 0.4875, "step": 3104 }, { "epoch": 2.3632210165619645, "grad_norm": 41.30006790161133, "learning_rate": 1.1009284705976453e-06, "loss": 0.7065, "step": 3105 }, { "epoch": 2.363982486198363, "grad_norm": 38.74078369140625, "learning_rate": 1.0984031850363885e-06, "loss": 0.813, "step": 3106 }, { "epoch": 2.364743955834761, "grad_norm": 37.61562728881836, "learning_rate": 1.09588044159747e-06, "loss": 0.6874, "step": 3107 }, { "epoch": 2.3655054254711594, "grad_norm": 30.17512321472168, "learning_rate": 1.093360241924603e-06, "loss": 0.6266, "step": 3108 }, { "epoch": 2.3662668951075574, "grad_norm": 42.72329330444336, "learning_rate": 1.0908425876598512e-06, "loss": 0.643, "step": 3109 }, { "epoch": 2.367028364743956, "grad_norm": 32.57030487060547, "learning_rate": 1.0883274804436155e-06, "loss": 0.5939, "step": 3110 }, { "epoch": 2.3677898343803543, "grad_norm": 31.383304595947266, "learning_rate": 1.0858149219146368e-06, "loss": 0.6572, "step": 3111 }, { "epoch": 2.3685513040167523, "grad_norm": 58.620025634765625, "learning_rate": 1.083304913709996e-06, "loss": 0.6268, "step": 3112 }, { "epoch": 2.3693127736531507, "grad_norm": 47.88822555541992, "learning_rate": 1.0807974574651154e-06, "loss": 0.8299, "step": 3113 }, { "epoch": 2.3700742432895487, "grad_norm": 29.161052703857422, "learning_rate": 1.0782925548137496e-06, "loss": 0.5904, "step": 3114 }, { "epoch": 2.370835712925947, "grad_norm": 39.694984436035156, "learning_rate": 1.0757902073879933e-06, "loss": 0.6844, "step": 3115 }, { "epoch": 2.371597182562345, "grad_norm": 46.310211181640625, "learning_rate": 1.0732904168182712e-06, "loss": 0.661, "step": 3116 }, { "epoch": 2.3723586521987436, "grad_norm": 38.949119567871094, "learning_rate": 1.0707931847333487e-06, "loss": 0.8239, "step": 3117 }, { "epoch": 2.3731201218351416, "grad_norm": 37.00999450683594, "learning_rate": 1.0682985127603202e-06, "loss": 0.7594, "step": 3118 }, { "epoch": 2.37388159147154, "grad_norm": 49.75098419189453, "learning_rate": 1.0658064025246128e-06, "loss": 0.887, "step": 3119 }, { "epoch": 2.374643061107938, "grad_norm": 35.04350280761719, "learning_rate": 1.0633168556499846e-06, "loss": 0.5695, "step": 3120 }, { "epoch": 2.3754045307443366, "grad_norm": 37.704280853271484, "learning_rate": 1.0608298737585222e-06, "loss": 0.6707, "step": 3121 }, { "epoch": 2.376166000380735, "grad_norm": 42.48690414428711, "learning_rate": 1.0583454584706454e-06, "loss": 0.5975, "step": 3122 }, { "epoch": 2.376927470017133, "grad_norm": 34.39491653442383, "learning_rate": 1.0558636114050952e-06, "loss": 0.6821, "step": 3123 }, { "epoch": 2.3776889396535315, "grad_norm": 31.53750991821289, "learning_rate": 1.053384334178944e-06, "loss": 0.5742, "step": 3124 }, { "epoch": 2.3784504092899295, "grad_norm": 35.056297302246094, "learning_rate": 1.0509076284075864e-06, "loss": 0.6658, "step": 3125 }, { "epoch": 2.379211878926328, "grad_norm": 33.176475524902344, "learning_rate": 1.0484334957047465e-06, "loss": 0.6845, "step": 3126 }, { "epoch": 2.379973348562726, "grad_norm": 41.598907470703125, "learning_rate": 1.0459619376824675e-06, "loss": 0.8947, "step": 3127 }, { "epoch": 2.3807348181991244, "grad_norm": 38.73931121826172, "learning_rate": 1.0434929559511164e-06, "loss": 0.7035, "step": 3128 }, { "epoch": 2.381496287835523, "grad_norm": 31.191829681396484, "learning_rate": 1.0410265521193823e-06, "loss": 0.5187, "step": 3129 }, { "epoch": 2.382257757471921, "grad_norm": 34.671268463134766, "learning_rate": 1.0385627277942728e-06, "loss": 0.7162, "step": 3130 }, { "epoch": 2.383019227108319, "grad_norm": 27.04009437561035, "learning_rate": 1.036101484581117e-06, "loss": 0.5286, "step": 3131 }, { "epoch": 2.3837806967447173, "grad_norm": 38.101287841796875, "learning_rate": 1.03364282408356e-06, "loss": 0.6403, "step": 3132 }, { "epoch": 2.3845421663811157, "grad_norm": 31.56092071533203, "learning_rate": 1.0311867479035648e-06, "loss": 0.7044, "step": 3133 }, { "epoch": 2.3853036360175137, "grad_norm": 33.69940948486328, "learning_rate": 1.0287332576414138e-06, "loss": 0.6847, "step": 3134 }, { "epoch": 2.386065105653912, "grad_norm": 37.53281021118164, "learning_rate": 1.0262823548957007e-06, "loss": 0.7549, "step": 3135 }, { "epoch": 2.38682657529031, "grad_norm": 37.16306686401367, "learning_rate": 1.0238340412633336e-06, "loss": 0.775, "step": 3136 }, { "epoch": 2.3875880449267086, "grad_norm": 33.720008850097656, "learning_rate": 1.0213883183395351e-06, "loss": 0.6375, "step": 3137 }, { "epoch": 2.3883495145631066, "grad_norm": 37.69129943847656, "learning_rate": 1.0189451877178386e-06, "loss": 0.7327, "step": 3138 }, { "epoch": 2.389110984199505, "grad_norm": 34.87730026245117, "learning_rate": 1.0165046509900895e-06, "loss": 0.6448, "step": 3139 }, { "epoch": 2.3898724538359035, "grad_norm": 35.74290466308594, "learning_rate": 1.0140667097464419e-06, "loss": 0.6561, "step": 3140 }, { "epoch": 2.3906339234723015, "grad_norm": 31.25864028930664, "learning_rate": 1.0116313655753602e-06, "loss": 0.6319, "step": 3141 }, { "epoch": 2.3913953931087, "grad_norm": 35.94463348388672, "learning_rate": 1.0091986200636144e-06, "loss": 0.6927, "step": 3142 }, { "epoch": 2.392156862745098, "grad_norm": 38.08346939086914, "learning_rate": 1.0067684747962858e-06, "loss": 0.6094, "step": 3143 }, { "epoch": 2.3929183323814964, "grad_norm": 32.85715866088867, "learning_rate": 1.0043409313567565e-06, "loss": 0.6328, "step": 3144 }, { "epoch": 2.3936798020178944, "grad_norm": 40.80418395996094, "learning_rate": 1.0019159913267156e-06, "loss": 0.6608, "step": 3145 }, { "epoch": 2.394441271654293, "grad_norm": 33.327640533447266, "learning_rate": 9.994936562861563e-07, "loss": 0.6599, "step": 3146 }, { "epoch": 2.395202741290691, "grad_norm": 37.75202941894531, "learning_rate": 9.970739278133735e-07, "loss": 0.8275, "step": 3147 }, { "epoch": 2.3959642109270893, "grad_norm": 27.203227996826172, "learning_rate": 9.946568074849644e-07, "loss": 0.5846, "step": 3148 }, { "epoch": 2.3967256805634873, "grad_norm": 34.14052200317383, "learning_rate": 9.92242296875826e-07, "loss": 0.7642, "step": 3149 }, { "epoch": 2.397487150199886, "grad_norm": 36.82248306274414, "learning_rate": 9.898303975591555e-07, "loss": 0.6688, "step": 3150 }, { "epoch": 2.3982486198362842, "grad_norm": 33.48533630371094, "learning_rate": 9.874211111064503e-07, "loss": 0.7809, "step": 3151 }, { "epoch": 2.3990100894726822, "grad_norm": 32.04188919067383, "learning_rate": 9.850144390875022e-07, "loss": 0.8859, "step": 3152 }, { "epoch": 2.3997715591090807, "grad_norm": 38.33725357055664, "learning_rate": 9.826103830704037e-07, "loss": 0.7527, "step": 3153 }, { "epoch": 2.4005330287454787, "grad_norm": 35.849525451660156, "learning_rate": 9.802089446215352e-07, "loss": 0.7079, "step": 3154 }, { "epoch": 2.401294498381877, "grad_norm": 51.39255905151367, "learning_rate": 9.77810125305581e-07, "loss": 0.6748, "step": 3155 }, { "epoch": 2.402055968018275, "grad_norm": 29.956409454345703, "learning_rate": 9.754139266855134e-07, "loss": 0.6596, "step": 3156 }, { "epoch": 2.4028174376546736, "grad_norm": 37.335269927978516, "learning_rate": 9.730203503225982e-07, "loss": 0.7338, "step": 3157 }, { "epoch": 2.4035789072910716, "grad_norm": 36.30415725708008, "learning_rate": 9.706293977763926e-07, "loss": 0.6629, "step": 3158 }, { "epoch": 2.40434037692747, "grad_norm": 34.51976776123047, "learning_rate": 9.68241070604743e-07, "loss": 0.7728, "step": 3159 }, { "epoch": 2.405101846563868, "grad_norm": 35.01738739013672, "learning_rate": 9.658553703637897e-07, "loss": 0.754, "step": 3160 }, { "epoch": 2.4058633162002665, "grad_norm": 31.475109100341797, "learning_rate": 9.634722986079575e-07, "loss": 0.6711, "step": 3161 }, { "epoch": 2.406624785836665, "grad_norm": 32.53989028930664, "learning_rate": 9.610918568899569e-07, "loss": 0.5546, "step": 3162 }, { "epoch": 2.407386255473063, "grad_norm": 30.577043533325195, "learning_rate": 9.58714046760788e-07, "loss": 0.6493, "step": 3163 }, { "epoch": 2.4081477251094614, "grad_norm": 27.27184295654297, "learning_rate": 9.563388697697373e-07, "loss": 0.5575, "step": 3164 }, { "epoch": 2.4089091947458594, "grad_norm": 37.577518463134766, "learning_rate": 9.53966327464373e-07, "loss": 0.6048, "step": 3165 }, { "epoch": 2.409670664382258, "grad_norm": 44.76778030395508, "learning_rate": 9.51596421390547e-07, "loss": 0.8068, "step": 3166 }, { "epoch": 2.410432134018656, "grad_norm": 35.18150329589844, "learning_rate": 9.49229153092393e-07, "loss": 0.7073, "step": 3167 }, { "epoch": 2.4111936036550543, "grad_norm": 29.291017532348633, "learning_rate": 9.468645241123303e-07, "loss": 0.5892, "step": 3168 }, { "epoch": 2.4119550732914523, "grad_norm": 35.25429916381836, "learning_rate": 9.445025359910543e-07, "loss": 0.6258, "step": 3169 }, { "epoch": 2.4127165429278508, "grad_norm": 41.41679382324219, "learning_rate": 9.421431902675388e-07, "loss": 0.6249, "step": 3170 }, { "epoch": 2.4134780125642488, "grad_norm": 36.7808952331543, "learning_rate": 9.397864884790369e-07, "loss": 0.6191, "step": 3171 }, { "epoch": 2.414239482200647, "grad_norm": 40.87490463256836, "learning_rate": 9.374324321610839e-07, "loss": 0.7523, "step": 3172 }, { "epoch": 2.4150009518370457, "grad_norm": 34.91834259033203, "learning_rate": 9.350810228474855e-07, "loss": 0.6626, "step": 3173 }, { "epoch": 2.4157624214734437, "grad_norm": 40.24203109741211, "learning_rate": 9.327322620703244e-07, "loss": 0.7419, "step": 3174 }, { "epoch": 2.416523891109842, "grad_norm": 35.86165237426758, "learning_rate": 9.303861513599588e-07, "loss": 0.678, "step": 3175 }, { "epoch": 2.41728536074624, "grad_norm": 25.342987060546875, "learning_rate": 9.28042692245017e-07, "loss": 0.5896, "step": 3176 }, { "epoch": 2.4180468303826386, "grad_norm": 36.27814865112305, "learning_rate": 9.257018862524064e-07, "loss": 0.7016, "step": 3177 }, { "epoch": 2.4188083000190366, "grad_norm": 27.966636657714844, "learning_rate": 9.233637349072971e-07, "loss": 0.725, "step": 3178 }, { "epoch": 2.419569769655435, "grad_norm": 27.772777557373047, "learning_rate": 9.210282397331355e-07, "loss": 0.56, "step": 3179 }, { "epoch": 2.4203312392918335, "grad_norm": 45.875980377197266, "learning_rate": 9.186954022516343e-07, "loss": 0.7568, "step": 3180 }, { "epoch": 2.4210927089282315, "grad_norm": 36.39308166503906, "learning_rate": 9.163652239827775e-07, "loss": 0.7316, "step": 3181 }, { "epoch": 2.42185417856463, "grad_norm": 37.025997161865234, "learning_rate": 9.140377064448147e-07, "loss": 0.7751, "step": 3182 }, { "epoch": 2.422615648201028, "grad_norm": 32.349632263183594, "learning_rate": 9.117128511542617e-07, "loss": 0.683, "step": 3183 }, { "epoch": 2.4233771178374264, "grad_norm": 34.14254379272461, "learning_rate": 9.093906596258995e-07, "loss": 0.7436, "step": 3184 }, { "epoch": 2.4241385874738244, "grad_norm": 28.53260040283203, "learning_rate": 9.070711333727744e-07, "loss": 0.5518, "step": 3185 }, { "epoch": 2.424900057110223, "grad_norm": 38.57051086425781, "learning_rate": 9.047542739061954e-07, "loss": 0.7201, "step": 3186 }, { "epoch": 2.425661526746621, "grad_norm": 31.446369171142578, "learning_rate": 9.024400827357344e-07, "loss": 0.6664, "step": 3187 }, { "epoch": 2.4264229963830193, "grad_norm": 33.91133499145508, "learning_rate": 9.001285613692235e-07, "loss": 0.6408, "step": 3188 }, { "epoch": 2.4271844660194173, "grad_norm": 35.977447509765625, "learning_rate": 8.97819711312758e-07, "loss": 0.8747, "step": 3189 }, { "epoch": 2.4279459356558157, "grad_norm": 32.477943420410156, "learning_rate": 8.95513534070691e-07, "loss": 0.758, "step": 3190 }, { "epoch": 2.428707405292214, "grad_norm": 31.003774642944336, "learning_rate": 8.932100311456332e-07, "loss": 0.5841, "step": 3191 }, { "epoch": 2.429468874928612, "grad_norm": 32.30789566040039, "learning_rate": 8.909092040384538e-07, "loss": 0.5312, "step": 3192 }, { "epoch": 2.4302303445650106, "grad_norm": 32.88601303100586, "learning_rate": 8.886110542482789e-07, "loss": 0.6425, "step": 3193 }, { "epoch": 2.4309918142014086, "grad_norm": 44.36073303222656, "learning_rate": 8.863155832724895e-07, "loss": 0.7619, "step": 3194 }, { "epoch": 2.431753283837807, "grad_norm": 35.60179901123047, "learning_rate": 8.840227926067213e-07, "loss": 0.6674, "step": 3195 }, { "epoch": 2.432514753474205, "grad_norm": 28.171323776245117, "learning_rate": 8.817326837448642e-07, "loss": 0.5779, "step": 3196 }, { "epoch": 2.4332762231106035, "grad_norm": 37.266204833984375, "learning_rate": 8.794452581790586e-07, "loss": 0.856, "step": 3197 }, { "epoch": 2.4340376927470015, "grad_norm": 36.53726577758789, "learning_rate": 8.771605173997011e-07, "loss": 0.6753, "step": 3198 }, { "epoch": 2.4347991623834, "grad_norm": 43.94609832763672, "learning_rate": 8.748784628954348e-07, "loss": 0.702, "step": 3199 }, { "epoch": 2.435560632019798, "grad_norm": 37.44776916503906, "learning_rate": 8.72599096153155e-07, "loss": 0.7431, "step": 3200 }, { "epoch": 2.4363221016561964, "grad_norm": 38.80601119995117, "learning_rate": 8.703224186580012e-07, "loss": 0.7403, "step": 3201 }, { "epoch": 2.437083571292595, "grad_norm": 40.295692443847656, "learning_rate": 8.68048431893368e-07, "loss": 0.8185, "step": 3202 }, { "epoch": 2.437845040928993, "grad_norm": 36.400535583496094, "learning_rate": 8.657771373408913e-07, "loss": 0.7546, "step": 3203 }, { "epoch": 2.4386065105653913, "grad_norm": 37.648712158203125, "learning_rate": 8.635085364804547e-07, "loss": 0.8406, "step": 3204 }, { "epoch": 2.4393679802017894, "grad_norm": 38.66661071777344, "learning_rate": 8.612426307901856e-07, "loss": 0.8269, "step": 3205 }, { "epoch": 2.440129449838188, "grad_norm": 35.063255310058594, "learning_rate": 8.589794217464586e-07, "loss": 0.7033, "step": 3206 }, { "epoch": 2.440890919474586, "grad_norm": 36.80625534057617, "learning_rate": 8.567189108238883e-07, "loss": 0.6908, "step": 3207 }, { "epoch": 2.4416523891109843, "grad_norm": 33.60029983520508, "learning_rate": 8.544610994953317e-07, "loss": 0.7691, "step": 3208 }, { "epoch": 2.4424138587473823, "grad_norm": 33.05067443847656, "learning_rate": 8.522059892318851e-07, "loss": 0.7365, "step": 3209 }, { "epoch": 2.4431753283837807, "grad_norm": 40.41891098022461, "learning_rate": 8.499535815028903e-07, "loss": 0.677, "step": 3210 }, { "epoch": 2.4439367980201787, "grad_norm": 43.49858856201172, "learning_rate": 8.477038777759228e-07, "loss": 0.92, "step": 3211 }, { "epoch": 2.444698267656577, "grad_norm": 33.42390441894531, "learning_rate": 8.454568795167989e-07, "loss": 0.7779, "step": 3212 }, { "epoch": 2.4454597372929756, "grad_norm": 44.392234802246094, "learning_rate": 8.432125881895714e-07, "loss": 0.8259, "step": 3213 }, { "epoch": 2.4462212069293736, "grad_norm": 44.65909194946289, "learning_rate": 8.409710052565289e-07, "loss": 0.6117, "step": 3214 }, { "epoch": 2.446982676565772, "grad_norm": 34.802978515625, "learning_rate": 8.387321321781977e-07, "loss": 0.549, "step": 3215 }, { "epoch": 2.44774414620217, "grad_norm": 31.227054595947266, "learning_rate": 8.364959704133374e-07, "loss": 0.5957, "step": 3216 }, { "epoch": 2.4485056158385685, "grad_norm": 38.64544677734375, "learning_rate": 8.342625214189376e-07, "loss": 0.802, "step": 3217 }, { "epoch": 2.4492670854749665, "grad_norm": 38.57969284057617, "learning_rate": 8.320317866502231e-07, "loss": 0.6781, "step": 3218 }, { "epoch": 2.450028555111365, "grad_norm": 77.03079223632812, "learning_rate": 8.298037675606524e-07, "loss": 0.7381, "step": 3219 }, { "epoch": 2.4507900247477634, "grad_norm": 38.130008697509766, "learning_rate": 8.27578465601912e-07, "loss": 0.6289, "step": 3220 }, { "epoch": 2.4515514943841614, "grad_norm": 32.493385314941406, "learning_rate": 8.253558822239177e-07, "loss": 0.6509, "step": 3221 }, { "epoch": 2.45231296402056, "grad_norm": 30.219741821289062, "learning_rate": 8.23136018874815e-07, "loss": 0.6186, "step": 3222 }, { "epoch": 2.453074433656958, "grad_norm": 37.805511474609375, "learning_rate": 8.209188770009769e-07, "loss": 0.6566, "step": 3223 }, { "epoch": 2.4538359032933563, "grad_norm": 34.184906005859375, "learning_rate": 8.187044580470033e-07, "loss": 0.758, "step": 3224 }, { "epoch": 2.4545973729297543, "grad_norm": 27.575090408325195, "learning_rate": 8.164927634557195e-07, "loss": 0.6348, "step": 3225 }, { "epoch": 2.4553588425661528, "grad_norm": 29.882844924926758, "learning_rate": 8.142837946681753e-07, "loss": 0.6274, "step": 3226 }, { "epoch": 2.456120312202551, "grad_norm": 50.78447341918945, "learning_rate": 8.120775531236475e-07, "loss": 0.9901, "step": 3227 }, { "epoch": 2.4568817818389492, "grad_norm": 43.497718811035156, "learning_rate": 8.098740402596317e-07, "loss": 0.7244, "step": 3228 }, { "epoch": 2.4576432514753472, "grad_norm": 35.53345489501953, "learning_rate": 8.07673257511849e-07, "loss": 0.7119, "step": 3229 }, { "epoch": 2.4584047211117457, "grad_norm": 30.82868194580078, "learning_rate": 8.054752063142391e-07, "loss": 0.6399, "step": 3230 }, { "epoch": 2.459166190748144, "grad_norm": 30.110212326049805, "learning_rate": 8.032798880989629e-07, "loss": 0.6575, "step": 3231 }, { "epoch": 2.459927660384542, "grad_norm": 34.45428466796875, "learning_rate": 8.010873042964018e-07, "loss": 0.5863, "step": 3232 }, { "epoch": 2.4606891300209406, "grad_norm": 28.33201026916504, "learning_rate": 7.988974563351543e-07, "loss": 0.6042, "step": 3233 }, { "epoch": 2.4614505996573386, "grad_norm": 39.134918212890625, "learning_rate": 7.967103456420355e-07, "loss": 0.8446, "step": 3234 }, { "epoch": 2.462212069293737, "grad_norm": 38.32954406738281, "learning_rate": 7.945259736420785e-07, "loss": 0.7141, "step": 3235 }, { "epoch": 2.462973538930135, "grad_norm": 38.261474609375, "learning_rate": 7.923443417585324e-07, "loss": 0.6845, "step": 3236 }, { "epoch": 2.4637350085665335, "grad_norm": 33.66992950439453, "learning_rate": 7.901654514128598e-07, "loss": 0.7422, "step": 3237 }, { "epoch": 2.4644964782029315, "grad_norm": 30.132701873779297, "learning_rate": 7.879893040247371e-07, "loss": 0.44, "step": 3238 }, { "epoch": 2.46525794783933, "grad_norm": 40.65178680419922, "learning_rate": 7.858159010120542e-07, "loss": 0.893, "step": 3239 }, { "epoch": 2.466019417475728, "grad_norm": 30.339778900146484, "learning_rate": 7.836452437909114e-07, "loss": 0.6057, "step": 3240 }, { "epoch": 2.4667808871121264, "grad_norm": 27.549699783325195, "learning_rate": 7.814773337756215e-07, "loss": 0.5334, "step": 3241 }, { "epoch": 2.467542356748525, "grad_norm": 38.136756896972656, "learning_rate": 7.793121723787073e-07, "loss": 0.7416, "step": 3242 }, { "epoch": 2.468303826384923, "grad_norm": 46.44742965698242, "learning_rate": 7.771497610108981e-07, "loss": 0.7563, "step": 3243 }, { "epoch": 2.4690652960213213, "grad_norm": 30.467182159423828, "learning_rate": 7.749901010811367e-07, "loss": 0.6459, "step": 3244 }, { "epoch": 2.4698267656577193, "grad_norm": 35.04523468017578, "learning_rate": 7.728331939965678e-07, "loss": 0.5723, "step": 3245 }, { "epoch": 2.4705882352941178, "grad_norm": 48.59148406982422, "learning_rate": 7.706790411625459e-07, "loss": 0.9678, "step": 3246 }, { "epoch": 2.4713497049305158, "grad_norm": 28.253170013427734, "learning_rate": 7.685276439826284e-07, "loss": 0.601, "step": 3247 }, { "epoch": 2.472111174566914, "grad_norm": 32.8360595703125, "learning_rate": 7.663790038585794e-07, "loss": 0.7336, "step": 3248 }, { "epoch": 2.472872644203312, "grad_norm": 36.011234283447266, "learning_rate": 7.642331221903654e-07, "loss": 0.8164, "step": 3249 }, { "epoch": 2.4736341138397107, "grad_norm": 35.2481803894043, "learning_rate": 7.620900003761561e-07, "loss": 0.7069, "step": 3250 }, { "epoch": 2.4743955834761087, "grad_norm": 34.62240982055664, "learning_rate": 7.599496398123224e-07, "loss": 0.6437, "step": 3251 }, { "epoch": 2.475157053112507, "grad_norm": 34.086978912353516, "learning_rate": 7.578120418934353e-07, "loss": 0.5735, "step": 3252 }, { "epoch": 2.4759185227489056, "grad_norm": 33.89104461669922, "learning_rate": 7.556772080122698e-07, "loss": 0.9353, "step": 3253 }, { "epoch": 2.4766799923853036, "grad_norm": 37.60442352294922, "learning_rate": 7.535451395597954e-07, "loss": 0.6807, "step": 3254 }, { "epoch": 2.477441462021702, "grad_norm": 35.74176788330078, "learning_rate": 7.514158379251829e-07, "loss": 0.6626, "step": 3255 }, { "epoch": 2.4782029316581, "grad_norm": 34.02021789550781, "learning_rate": 7.49289304495796e-07, "loss": 0.6773, "step": 3256 }, { "epoch": 2.4789644012944985, "grad_norm": 41.91249465942383, "learning_rate": 7.471655406572003e-07, "loss": 0.69, "step": 3257 }, { "epoch": 2.4797258709308965, "grad_norm": 28.345552444458008, "learning_rate": 7.450445477931534e-07, "loss": 0.6331, "step": 3258 }, { "epoch": 2.480487340567295, "grad_norm": 29.161529541015625, "learning_rate": 7.429263272856085e-07, "loss": 0.7066, "step": 3259 }, { "epoch": 2.4812488102036934, "grad_norm": 31.47775650024414, "learning_rate": 7.408108805147118e-07, "loss": 0.6148, "step": 3260 }, { "epoch": 2.4820102798400914, "grad_norm": 27.9711971282959, "learning_rate": 7.386982088588018e-07, "loss": 0.7034, "step": 3261 }, { "epoch": 2.4827717494764894, "grad_norm": 40.66910934448242, "learning_rate": 7.36588313694413e-07, "loss": 0.5681, "step": 3262 }, { "epoch": 2.483533219112888, "grad_norm": 38.96367645263672, "learning_rate": 7.344811963962645e-07, "loss": 0.7525, "step": 3263 }, { "epoch": 2.4842946887492863, "grad_norm": 29.342947006225586, "learning_rate": 7.323768583372675e-07, "loss": 0.5684, "step": 3264 }, { "epoch": 2.4850561583856843, "grad_norm": 30.188907623291016, "learning_rate": 7.302753008885272e-07, "loss": 0.574, "step": 3265 }, { "epoch": 2.4858176280220827, "grad_norm": 34.38759231567383, "learning_rate": 7.281765254193312e-07, "loss": 0.8253, "step": 3266 }, { "epoch": 2.4865790976584807, "grad_norm": 30.67691993713379, "learning_rate": 7.260805332971566e-07, "loss": 0.4478, "step": 3267 }, { "epoch": 2.487340567294879, "grad_norm": 32.5081787109375, "learning_rate": 7.239873258876668e-07, "loss": 0.6232, "step": 3268 }, { "epoch": 2.488102036931277, "grad_norm": 43.99415969848633, "learning_rate": 7.218969045547097e-07, "loss": 0.7726, "step": 3269 }, { "epoch": 2.4888635065676756, "grad_norm": 30.899072647094727, "learning_rate": 7.198092706603232e-07, "loss": 0.6715, "step": 3270 }, { "epoch": 2.489624976204074, "grad_norm": 29.64040184020996, "learning_rate": 7.177244255647209e-07, "loss": 0.7098, "step": 3271 }, { "epoch": 2.490386445840472, "grad_norm": 36.41777420043945, "learning_rate": 7.156423706263044e-07, "loss": 0.8043, "step": 3272 }, { "epoch": 2.4911479154768705, "grad_norm": 31.841764450073242, "learning_rate": 7.135631072016558e-07, "loss": 0.5845, "step": 3273 }, { "epoch": 2.4919093851132685, "grad_norm": 31.944303512573242, "learning_rate": 7.114866366455403e-07, "loss": 0.606, "step": 3274 }, { "epoch": 2.492670854749667, "grad_norm": 32.41287612915039, "learning_rate": 7.094129603109018e-07, "loss": 0.5937, "step": 3275 }, { "epoch": 2.493432324386065, "grad_norm": 35.971961975097656, "learning_rate": 7.073420795488628e-07, "loss": 0.708, "step": 3276 }, { "epoch": 2.4941937940224634, "grad_norm": 34.15237045288086, "learning_rate": 7.052739957087257e-07, "loss": 0.7569, "step": 3277 }, { "epoch": 2.4949552636588614, "grad_norm": 30.546775817871094, "learning_rate": 7.032087101379703e-07, "loss": 0.5926, "step": 3278 }, { "epoch": 2.49571673329526, "grad_norm": 29.190122604370117, "learning_rate": 7.011462241822525e-07, "loss": 0.7622, "step": 3279 }, { "epoch": 2.496478202931658, "grad_norm": 39.75563049316406, "learning_rate": 6.990865391854046e-07, "loss": 0.6862, "step": 3280 }, { "epoch": 2.4972396725680563, "grad_norm": 37.38508224487305, "learning_rate": 6.97029656489433e-07, "loss": 0.7362, "step": 3281 }, { "epoch": 2.498001142204455, "grad_norm": 31.408437728881836, "learning_rate": 6.949755774345218e-07, "loss": 0.6642, "step": 3282 }, { "epoch": 2.498762611840853, "grad_norm": 33.157936096191406, "learning_rate": 6.929243033590233e-07, "loss": 0.6608, "step": 3283 }, { "epoch": 2.4995240814772512, "grad_norm": 36.63829803466797, "learning_rate": 6.908758355994655e-07, "loss": 0.6124, "step": 3284 }, { "epoch": 2.5002855511136493, "grad_norm": 25.74787139892578, "learning_rate": 6.888301754905469e-07, "loss": 0.6168, "step": 3285 }, { "epoch": 2.5010470207500477, "grad_norm": 34.36436080932617, "learning_rate": 6.867873243651363e-07, "loss": 0.6323, "step": 3286 }, { "epoch": 2.5018084903864457, "grad_norm": 26.19143295288086, "learning_rate": 6.847472835542723e-07, "loss": 0.4868, "step": 3287 }, { "epoch": 2.502569960022844, "grad_norm": 38.313175201416016, "learning_rate": 6.82710054387164e-07, "loss": 0.6772, "step": 3288 }, { "epoch": 2.5033314296592426, "grad_norm": 35.459468841552734, "learning_rate": 6.806756381911867e-07, "loss": 0.7122, "step": 3289 }, { "epoch": 2.5040928992956406, "grad_norm": 32.22385025024414, "learning_rate": 6.786440362918823e-07, "loss": 0.6512, "step": 3290 }, { "epoch": 2.5048543689320386, "grad_norm": 33.415191650390625, "learning_rate": 6.76615250012963e-07, "loss": 0.7123, "step": 3291 }, { "epoch": 2.505615838568437, "grad_norm": 30.48807144165039, "learning_rate": 6.745892806763015e-07, "loss": 0.5247, "step": 3292 }, { "epoch": 2.5063773082048355, "grad_norm": 44.25031661987305, "learning_rate": 6.725661296019387e-07, "loss": 0.7984, "step": 3293 }, { "epoch": 2.5071387778412335, "grad_norm": 36.04548645019531, "learning_rate": 6.705457981080765e-07, "loss": 0.6049, "step": 3294 }, { "epoch": 2.507900247477632, "grad_norm": 39.66203308105469, "learning_rate": 6.685282875110822e-07, "loss": 0.6781, "step": 3295 }, { "epoch": 2.50866171711403, "grad_norm": 33.98748016357422, "learning_rate": 6.665135991254828e-07, "loss": 0.4998, "step": 3296 }, { "epoch": 2.5094231867504284, "grad_norm": 37.41563415527344, "learning_rate": 6.645017342639687e-07, "loss": 0.6471, "step": 3297 }, { "epoch": 2.5101846563868264, "grad_norm": 32.6900749206543, "learning_rate": 6.624926942373877e-07, "loss": 0.8241, "step": 3298 }, { "epoch": 2.510946126023225, "grad_norm": 29.849157333374023, "learning_rate": 6.604864803547511e-07, "loss": 0.5766, "step": 3299 }, { "epoch": 2.5117075956596233, "grad_norm": 35.249717712402344, "learning_rate": 6.584830939232256e-07, "loss": 0.7146, "step": 3300 }, { "epoch": 2.5124690652960213, "grad_norm": 36.48558044433594, "learning_rate": 6.564825362481375e-07, "loss": 0.6646, "step": 3301 }, { "epoch": 2.5132305349324193, "grad_norm": 37.159576416015625, "learning_rate": 6.544848086329663e-07, "loss": 0.8065, "step": 3302 }, { "epoch": 2.5139920045688178, "grad_norm": 35.96522521972656, "learning_rate": 6.524899123793532e-07, "loss": 0.7403, "step": 3303 }, { "epoch": 2.514753474205216, "grad_norm": 36.514373779296875, "learning_rate": 6.504978487870905e-07, "loss": 0.7085, "step": 3304 }, { "epoch": 2.5155149438416142, "grad_norm": 31.492340087890625, "learning_rate": 6.485086191541263e-07, "loss": 0.6327, "step": 3305 }, { "epoch": 2.5162764134780127, "grad_norm": 36.401004791259766, "learning_rate": 6.465222247765618e-07, "loss": 0.5807, "step": 3306 }, { "epoch": 2.5170378831144107, "grad_norm": 32.130821228027344, "learning_rate": 6.445386669486503e-07, "loss": 0.5328, "step": 3307 }, { "epoch": 2.517799352750809, "grad_norm": 45.19773864746094, "learning_rate": 6.425579469628001e-07, "loss": 0.751, "step": 3308 }, { "epoch": 2.518560822387207, "grad_norm": 36.46647644042969, "learning_rate": 6.405800661095674e-07, "loss": 0.6524, "step": 3309 }, { "epoch": 2.5193222920236056, "grad_norm": 28.64704132080078, "learning_rate": 6.386050256776571e-07, "loss": 0.6162, "step": 3310 }, { "epoch": 2.520083761660004, "grad_norm": 35.793235778808594, "learning_rate": 6.366328269539262e-07, "loss": 0.8826, "step": 3311 }, { "epoch": 2.520845231296402, "grad_norm": 40.13577651977539, "learning_rate": 6.346634712233807e-07, "loss": 0.5898, "step": 3312 }, { "epoch": 2.5216067009328, "grad_norm": 28.91039276123047, "learning_rate": 6.326969597691724e-07, "loss": 0.5582, "step": 3313 }, { "epoch": 2.5223681705691985, "grad_norm": 40.81425857543945, "learning_rate": 6.307332938726002e-07, "loss": 0.6107, "step": 3314 }, { "epoch": 2.523129640205597, "grad_norm": 35.9538459777832, "learning_rate": 6.287724748131096e-07, "loss": 0.8146, "step": 3315 }, { "epoch": 2.523891109841995, "grad_norm": 39.14362335205078, "learning_rate": 6.268145038682893e-07, "loss": 0.6157, "step": 3316 }, { "epoch": 2.5246525794783934, "grad_norm": 29.912878036499023, "learning_rate": 6.248593823138771e-07, "loss": 0.6505, "step": 3317 }, { "epoch": 2.5254140491147914, "grad_norm": 34.78255844116211, "learning_rate": 6.229071114237484e-07, "loss": 0.7132, "step": 3318 }, { "epoch": 2.52617551875119, "grad_norm": 31.73259925842285, "learning_rate": 6.209576924699224e-07, "loss": 0.5483, "step": 3319 }, { "epoch": 2.526936988387588, "grad_norm": 40.36781692504883, "learning_rate": 6.190111267225651e-07, "loss": 0.6954, "step": 3320 }, { "epoch": 2.5276984580239863, "grad_norm": 36.2351188659668, "learning_rate": 6.170674154499778e-07, "loss": 0.6242, "step": 3321 }, { "epoch": 2.5284599276603847, "grad_norm": 27.6554012298584, "learning_rate": 6.151265599186046e-07, "loss": 0.5851, "step": 3322 }, { "epoch": 2.5292213972967827, "grad_norm": 39.63099670410156, "learning_rate": 6.131885613930278e-07, "loss": 0.6121, "step": 3323 }, { "epoch": 2.529982866933181, "grad_norm": 29.502138137817383, "learning_rate": 6.112534211359672e-07, "loss": 0.5993, "step": 3324 }, { "epoch": 2.530744336569579, "grad_norm": 34.463531494140625, "learning_rate": 6.093211404082866e-07, "loss": 0.7216, "step": 3325 }, { "epoch": 2.5315058062059776, "grad_norm": 31.90226936340332, "learning_rate": 6.073917204689767e-07, "loss": 0.7217, "step": 3326 }, { "epoch": 2.5322672758423757, "grad_norm": 42.047664642333984, "learning_rate": 6.054651625751717e-07, "loss": 0.6002, "step": 3327 }, { "epoch": 2.533028745478774, "grad_norm": 31.268917083740234, "learning_rate": 6.035414679821372e-07, "loss": 0.4621, "step": 3328 }, { "epoch": 2.5337902151151726, "grad_norm": 32.0120735168457, "learning_rate": 6.01620637943276e-07, "loss": 0.6896, "step": 3329 }, { "epoch": 2.5345516847515706, "grad_norm": 32.8475456237793, "learning_rate": 5.997026737101236e-07, "loss": 0.7236, "step": 3330 }, { "epoch": 2.5353131543879686, "grad_norm": 36.36095428466797, "learning_rate": 5.97787576532346e-07, "loss": 0.7239, "step": 3331 }, { "epoch": 2.536074624024367, "grad_norm": 38.161258697509766, "learning_rate": 5.958753476577445e-07, "loss": 0.6683, "step": 3332 }, { "epoch": 2.5368360936607655, "grad_norm": 32.651817321777344, "learning_rate": 5.939659883322485e-07, "loss": 0.8134, "step": 3333 }, { "epoch": 2.5375975632971635, "grad_norm": 36.6313362121582, "learning_rate": 5.920594997999202e-07, "loss": 0.8373, "step": 3334 }, { "epoch": 2.538359032933562, "grad_norm": 25.140090942382812, "learning_rate": 5.901558833029497e-07, "loss": 0.5588, "step": 3335 }, { "epoch": 2.53912050256996, "grad_norm": 40.98360824584961, "learning_rate": 5.882551400816555e-07, "loss": 0.6653, "step": 3336 }, { "epoch": 2.5398819722063584, "grad_norm": 45.90095138549805, "learning_rate": 5.863572713744869e-07, "loss": 0.8036, "step": 3337 }, { "epoch": 2.5406434418427564, "grad_norm": 39.41801834106445, "learning_rate": 5.84462278418017e-07, "loss": 0.79, "step": 3338 }, { "epoch": 2.541404911479155, "grad_norm": 38.080482482910156, "learning_rate": 5.825701624469465e-07, "loss": 0.648, "step": 3339 }, { "epoch": 2.5421663811155533, "grad_norm": 38.65947341918945, "learning_rate": 5.806809246941019e-07, "loss": 0.6817, "step": 3340 }, { "epoch": 2.5429278507519513, "grad_norm": 31.653825759887695, "learning_rate": 5.787945663904332e-07, "loss": 0.6152, "step": 3341 }, { "epoch": 2.5436893203883493, "grad_norm": 33.19921112060547, "learning_rate": 5.769110887650159e-07, "loss": 0.6874, "step": 3342 }, { "epoch": 2.5444507900247477, "grad_norm": 42.28778076171875, "learning_rate": 5.750304930450468e-07, "loss": 0.7451, "step": 3343 }, { "epoch": 2.545212259661146, "grad_norm": 37.91328811645508, "learning_rate": 5.731527804558457e-07, "loss": 0.9217, "step": 3344 }, { "epoch": 2.545973729297544, "grad_norm": 30.38368034362793, "learning_rate": 5.712779522208544e-07, "loss": 0.6365, "step": 3345 }, { "epoch": 2.5467351989339426, "grad_norm": 39.84670639038086, "learning_rate": 5.694060095616355e-07, "loss": 0.5419, "step": 3346 }, { "epoch": 2.5474966685703406, "grad_norm": 29.445838928222656, "learning_rate": 5.675369536978708e-07, "loss": 0.7127, "step": 3347 }, { "epoch": 2.548258138206739, "grad_norm": 72.11616516113281, "learning_rate": 5.65670785847362e-07, "loss": 0.8324, "step": 3348 }, { "epoch": 2.549019607843137, "grad_norm": 37.63380813598633, "learning_rate": 5.638075072260257e-07, "loss": 0.7643, "step": 3349 }, { "epoch": 2.5497810774795355, "grad_norm": 34.307151794433594, "learning_rate": 5.619471190479014e-07, "loss": 0.6211, "step": 3350 }, { "epoch": 2.550542547115934, "grad_norm": 37.820594787597656, "learning_rate": 5.600896225251418e-07, "loss": 0.862, "step": 3351 }, { "epoch": 2.551304016752332, "grad_norm": 26.008636474609375, "learning_rate": 5.582350188680164e-07, "loss": 0.3206, "step": 3352 }, { "epoch": 2.55206548638873, "grad_norm": 30.15053939819336, "learning_rate": 5.563833092849092e-07, "loss": 0.5835, "step": 3353 }, { "epoch": 2.5528269560251284, "grad_norm": 39.12106704711914, "learning_rate": 5.54534494982319e-07, "loss": 0.7366, "step": 3354 }, { "epoch": 2.553588425661527, "grad_norm": 31.75062370300293, "learning_rate": 5.526885771648599e-07, "loss": 0.5522, "step": 3355 }, { "epoch": 2.554349895297925, "grad_norm": 29.356569290161133, "learning_rate": 5.508455570352567e-07, "loss": 0.6856, "step": 3356 }, { "epoch": 2.5551113649343233, "grad_norm": 26.322778701782227, "learning_rate": 5.490054357943442e-07, "loss": 0.6108, "step": 3357 }, { "epoch": 2.5558728345707213, "grad_norm": 38.73981475830078, "learning_rate": 5.47168214641074e-07, "loss": 0.7775, "step": 3358 }, { "epoch": 2.55663430420712, "grad_norm": 29.391321182250977, "learning_rate": 5.453338947725029e-07, "loss": 0.6548, "step": 3359 }, { "epoch": 2.557395773843518, "grad_norm": 28.950719833374023, "learning_rate": 5.435024773838004e-07, "loss": 0.6919, "step": 3360 }, { "epoch": 2.5581572434799162, "grad_norm": 28.83879852294922, "learning_rate": 5.416739636682433e-07, "loss": 0.6572, "step": 3361 }, { "epoch": 2.5589187131163147, "grad_norm": 29.00374984741211, "learning_rate": 5.398483548172162e-07, "loss": 0.4092, "step": 3362 }, { "epoch": 2.5596801827527127, "grad_norm": 32.884376525878906, "learning_rate": 5.380256520202138e-07, "loss": 0.5282, "step": 3363 }, { "epoch": 2.560441652389111, "grad_norm": 44.76089096069336, "learning_rate": 5.362058564648354e-07, "loss": 0.8774, "step": 3364 }, { "epoch": 2.561203122025509, "grad_norm": 33.45880889892578, "learning_rate": 5.343889693367837e-07, "loss": 0.7513, "step": 3365 }, { "epoch": 2.5619645916619076, "grad_norm": 30.845449447631836, "learning_rate": 5.325749918198691e-07, "loss": 0.7599, "step": 3366 }, { "epoch": 2.5627260612983056, "grad_norm": 36.88923263549805, "learning_rate": 5.307639250960079e-07, "loss": 0.741, "step": 3367 }, { "epoch": 2.563487530934704, "grad_norm": 34.100563049316406, "learning_rate": 5.289557703452164e-07, "loss": 0.5609, "step": 3368 }, { "epoch": 2.5642490005711025, "grad_norm": 35.31626510620117, "learning_rate": 5.271505287456153e-07, "loss": 0.6904, "step": 3369 }, { "epoch": 2.5650104702075005, "grad_norm": 35.07408142089844, "learning_rate": 5.253482014734268e-07, "loss": 0.7127, "step": 3370 }, { "epoch": 2.5657719398438985, "grad_norm": 34.95894241333008, "learning_rate": 5.235487897029734e-07, "loss": 0.6622, "step": 3371 }, { "epoch": 2.566533409480297, "grad_norm": 39.30010986328125, "learning_rate": 5.217522946066806e-07, "loss": 0.7727, "step": 3372 }, { "epoch": 2.5672948791166954, "grad_norm": 37.71702575683594, "learning_rate": 5.199587173550702e-07, "loss": 0.7094, "step": 3373 }, { "epoch": 2.5680563487530934, "grad_norm": 28.0982666015625, "learning_rate": 5.181680591167643e-07, "loss": 0.6277, "step": 3374 }, { "epoch": 2.568817818389492, "grad_norm": 38.60369873046875, "learning_rate": 5.163803210584845e-07, "loss": 0.6869, "step": 3375 }, { "epoch": 2.56957928802589, "grad_norm": 28.213590621948242, "learning_rate": 5.145955043450484e-07, "loss": 0.6334, "step": 3376 }, { "epoch": 2.5703407576622883, "grad_norm": 37.844757080078125, "learning_rate": 5.128136101393693e-07, "loss": 0.7721, "step": 3377 }, { "epoch": 2.5711022272986863, "grad_norm": 31.100242614746094, "learning_rate": 5.110346396024585e-07, "loss": 0.4746, "step": 3378 }, { "epoch": 2.5718636969350848, "grad_norm": 51.135066986083984, "learning_rate": 5.092585938934197e-07, "loss": 0.7608, "step": 3379 }, { "epoch": 2.572625166571483, "grad_norm": 34.93547821044922, "learning_rate": 5.074854741694529e-07, "loss": 0.6101, "step": 3380 }, { "epoch": 2.573386636207881, "grad_norm": 41.45198059082031, "learning_rate": 5.057152815858512e-07, "loss": 0.7433, "step": 3381 }, { "epoch": 2.574148105844279, "grad_norm": 44.33620071411133, "learning_rate": 5.039480172959999e-07, "loss": 0.8141, "step": 3382 }, { "epoch": 2.5749095754806777, "grad_norm": 32.7971076965332, "learning_rate": 5.021836824513759e-07, "loss": 0.7238, "step": 3383 }, { "epoch": 2.575671045117076, "grad_norm": 33.74524688720703, "learning_rate": 5.004222782015511e-07, "loss": 0.764, "step": 3384 }, { "epoch": 2.576432514753474, "grad_norm": 31.827407836914062, "learning_rate": 4.986638056941829e-07, "loss": 0.6094, "step": 3385 }, { "epoch": 2.5771939843898726, "grad_norm": 29.96603775024414, "learning_rate": 4.969082660750213e-07, "loss": 0.6681, "step": 3386 }, { "epoch": 2.5779554540262706, "grad_norm": 33.41102600097656, "learning_rate": 4.951556604879049e-07, "loss": 0.7018, "step": 3387 }, { "epoch": 2.578716923662669, "grad_norm": 36.09585952758789, "learning_rate": 4.934059900747596e-07, "loss": 0.8509, "step": 3388 }, { "epoch": 2.579478393299067, "grad_norm": 38.97381591796875, "learning_rate": 4.916592559756011e-07, "loss": 0.6506, "step": 3389 }, { "epoch": 2.5802398629354655, "grad_norm": 49.64519119262695, "learning_rate": 4.899154593285294e-07, "loss": 0.9509, "step": 3390 }, { "epoch": 2.581001332571864, "grad_norm": 33.01411437988281, "learning_rate": 4.88174601269732e-07, "loss": 0.5527, "step": 3391 }, { "epoch": 2.581762802208262, "grad_norm": 25.3448486328125, "learning_rate": 4.86436682933481e-07, "loss": 0.5289, "step": 3392 }, { "epoch": 2.58252427184466, "grad_norm": 33.98318862915039, "learning_rate": 4.847017054521347e-07, "loss": 0.6528, "step": 3393 }, { "epoch": 2.5832857414810584, "grad_norm": 33.410736083984375, "learning_rate": 4.829696699561337e-07, "loss": 0.7396, "step": 3394 }, { "epoch": 2.584047211117457, "grad_norm": 28.768924713134766, "learning_rate": 4.812405775740015e-07, "loss": 0.5864, "step": 3395 }, { "epoch": 2.584808680753855, "grad_norm": 42.3588981628418, "learning_rate": 4.795144294323456e-07, "loss": 0.7329, "step": 3396 }, { "epoch": 2.5855701503902533, "grad_norm": 49.22797775268555, "learning_rate": 4.777912266558532e-07, "loss": 0.7605, "step": 3397 }, { "epoch": 2.5863316200266513, "grad_norm": 32.233924865722656, "learning_rate": 4.760709703672939e-07, "loss": 0.7241, "step": 3398 }, { "epoch": 2.5870930896630497, "grad_norm": 41.359657287597656, "learning_rate": 4.7435366168751663e-07, "loss": 0.7001, "step": 3399 }, { "epoch": 2.5878545592994477, "grad_norm": 28.76319122314453, "learning_rate": 4.726393017354497e-07, "loss": 0.6079, "step": 3400 }, { "epoch": 2.588616028935846, "grad_norm": 48.84544372558594, "learning_rate": 4.709278916281018e-07, "loss": 1.0401, "step": 3401 }, { "epoch": 2.5893774985722446, "grad_norm": 35.713687896728516, "learning_rate": 4.6921943248055823e-07, "loss": 0.667, "step": 3402 }, { "epoch": 2.5901389682086426, "grad_norm": 32.111419677734375, "learning_rate": 4.6751392540598187e-07, "loss": 0.6938, "step": 3403 }, { "epoch": 2.590900437845041, "grad_norm": 35.14351272583008, "learning_rate": 4.658113715156098e-07, "loss": 0.7377, "step": 3404 }, { "epoch": 2.591661907481439, "grad_norm": 42.47444152832031, "learning_rate": 4.6411177191875934e-07, "loss": 0.7431, "step": 3405 }, { "epoch": 2.5924233771178375, "grad_norm": 39.425777435302734, "learning_rate": 4.6241512772282005e-07, "loss": 0.6805, "step": 3406 }, { "epoch": 2.5931848467542356, "grad_norm": 38.728614807128906, "learning_rate": 4.6072144003325657e-07, "loss": 0.8004, "step": 3407 }, { "epoch": 2.593946316390634, "grad_norm": 31.21794319152832, "learning_rate": 4.590307099536073e-07, "loss": 0.649, "step": 3408 }, { "epoch": 2.5947077860270324, "grad_norm": 30.840349197387695, "learning_rate": 4.5734293858548215e-07, "loss": 0.6155, "step": 3409 }, { "epoch": 2.5954692556634305, "grad_norm": 30.137075424194336, "learning_rate": 4.556581270285676e-07, "loss": 0.663, "step": 3410 }, { "epoch": 2.5962307252998285, "grad_norm": 28.837316513061523, "learning_rate": 4.5397627638061604e-07, "loss": 0.6216, "step": 3411 }, { "epoch": 2.596992194936227, "grad_norm": 36.52909851074219, "learning_rate": 4.52297387737452e-07, "loss": 0.6922, "step": 3412 }, { "epoch": 2.5977536645726254, "grad_norm": 32.10042953491211, "learning_rate": 4.5062146219297444e-07, "loss": 0.6035, "step": 3413 }, { "epoch": 2.5985151342090234, "grad_norm": 32.83551788330078, "learning_rate": 4.48948500839147e-07, "loss": 0.6132, "step": 3414 }, { "epoch": 2.599276603845422, "grad_norm": 46.814754486083984, "learning_rate": 4.4727850476600333e-07, "loss": 1.449, "step": 3415 }, { "epoch": 2.60003807348182, "grad_norm": 31.55084991455078, "learning_rate": 4.456114750616458e-07, "loss": 0.6027, "step": 3416 }, { "epoch": 2.6007995431182183, "grad_norm": 33.79789733886719, "learning_rate": 4.4394741281224174e-07, "loss": 0.7099, "step": 3417 }, { "epoch": 2.6015610127546163, "grad_norm": 36.500038146972656, "learning_rate": 4.422863191020299e-07, "loss": 0.6654, "step": 3418 }, { "epoch": 2.6023224823910147, "grad_norm": 36.461822509765625, "learning_rate": 4.4062819501330844e-07, "loss": 0.6128, "step": 3419 }, { "epoch": 2.603083952027413, "grad_norm": 31.262901306152344, "learning_rate": 4.389730416264454e-07, "loss": 0.5868, "step": 3420 }, { "epoch": 2.603845421663811, "grad_norm": 36.05514907836914, "learning_rate": 4.373208600198703e-07, "loss": 0.5706, "step": 3421 }, { "epoch": 2.604606891300209, "grad_norm": 32.31979751586914, "learning_rate": 4.3567165127007925e-07, "loss": 0.5873, "step": 3422 }, { "epoch": 2.6053683609366076, "grad_norm": 43.33320236206055, "learning_rate": 4.340254164516294e-07, "loss": 0.8517, "step": 3423 }, { "epoch": 2.606129830573006, "grad_norm": 45.7530517578125, "learning_rate": 4.3238215663714e-07, "loss": 0.8028, "step": 3424 }, { "epoch": 2.606891300209404, "grad_norm": 28.28118896484375, "learning_rate": 4.307418728972934e-07, "loss": 0.576, "step": 3425 }, { "epoch": 2.6076527698458025, "grad_norm": 31.27101707458496, "learning_rate": 4.2910456630083195e-07, "loss": 0.6714, "step": 3426 }, { "epoch": 2.6084142394822005, "grad_norm": 40.105804443359375, "learning_rate": 4.274702379145573e-07, "loss": 0.8184, "step": 3427 }, { "epoch": 2.609175709118599, "grad_norm": 32.44259262084961, "learning_rate": 4.2583888880333213e-07, "loss": 0.6147, "step": 3428 }, { "epoch": 2.609937178754997, "grad_norm": 42.042301177978516, "learning_rate": 4.242105200300767e-07, "loss": 0.7259, "step": 3429 }, { "epoch": 2.6106986483913954, "grad_norm": 36.14686584472656, "learning_rate": 4.2258513265577174e-07, "loss": 0.639, "step": 3430 }, { "epoch": 2.611460118027794, "grad_norm": 33.403202056884766, "learning_rate": 4.209627277394529e-07, "loss": 0.739, "step": 3431 }, { "epoch": 2.612221587664192, "grad_norm": 42.03188705444336, "learning_rate": 4.193433063382135e-07, "loss": 0.729, "step": 3432 }, { "epoch": 2.61298305730059, "grad_norm": 33.196754455566406, "learning_rate": 4.1772686950720285e-07, "loss": 0.6705, "step": 3433 }, { "epoch": 2.6137445269369883, "grad_norm": 34.554317474365234, "learning_rate": 4.161134182996257e-07, "loss": 0.7473, "step": 3434 }, { "epoch": 2.614505996573387, "grad_norm": 44.460750579833984, "learning_rate": 4.145029537667411e-07, "loss": 0.7542, "step": 3435 }, { "epoch": 2.615267466209785, "grad_norm": 31.848846435546875, "learning_rate": 4.12895476957863e-07, "loss": 0.6641, "step": 3436 }, { "epoch": 2.6160289358461832, "grad_norm": 30.39114761352539, "learning_rate": 4.112909889203576e-07, "loss": 0.707, "step": 3437 }, { "epoch": 2.6167904054825812, "grad_norm": 34.10219192504883, "learning_rate": 4.0968949069964346e-07, "loss": 0.6119, "step": 3438 }, { "epoch": 2.6175518751189797, "grad_norm": 30.824264526367188, "learning_rate": 4.080909833391944e-07, "loss": 0.6114, "step": 3439 }, { "epoch": 2.6183133447553777, "grad_norm": 30.517751693725586, "learning_rate": 4.0649546788053163e-07, "loss": 0.5474, "step": 3440 }, { "epoch": 2.619074814391776, "grad_norm": 36.72745895385742, "learning_rate": 4.049029453632286e-07, "loss": 0.8036, "step": 3441 }, { "epoch": 2.6198362840281746, "grad_norm": 34.19599151611328, "learning_rate": 4.03313416824907e-07, "loss": 0.6692, "step": 3442 }, { "epoch": 2.6205977536645726, "grad_norm": 34.96745300292969, "learning_rate": 4.017268833012411e-07, "loss": 0.6436, "step": 3443 }, { "epoch": 2.6213592233009706, "grad_norm": 38.9091682434082, "learning_rate": 4.0014334582595124e-07, "loss": 0.7779, "step": 3444 }, { "epoch": 2.622120692937369, "grad_norm": 36.63456344604492, "learning_rate": 3.985628054308055e-07, "loss": 0.6811, "step": 3445 }, { "epoch": 2.6228821625737675, "grad_norm": 43.78541946411133, "learning_rate": 3.9698526314562114e-07, "loss": 0.9605, "step": 3446 }, { "epoch": 2.6236436322101655, "grad_norm": 33.53181457519531, "learning_rate": 3.9541071999825933e-07, "loss": 0.7907, "step": 3447 }, { "epoch": 2.624405101846564, "grad_norm": 40.17458724975586, "learning_rate": 3.938391770146299e-07, "loss": 0.9693, "step": 3448 }, { "epoch": 2.625166571482962, "grad_norm": 35.14591979980469, "learning_rate": 3.9227063521868726e-07, "loss": 0.6043, "step": 3449 }, { "epoch": 2.6259280411193604, "grad_norm": 37.25705337524414, "learning_rate": 3.9070509563242654e-07, "loss": 0.7905, "step": 3450 }, { "epoch": 2.6266895107557584, "grad_norm": 37.913246154785156, "learning_rate": 3.891425592758924e-07, "loss": 0.529, "step": 3451 }, { "epoch": 2.627450980392157, "grad_norm": 40.3756103515625, "learning_rate": 3.875830271671699e-07, "loss": 0.7393, "step": 3452 }, { "epoch": 2.6282124500285553, "grad_norm": 30.89643096923828, "learning_rate": 3.8602650032238675e-07, "loss": 0.4992, "step": 3453 }, { "epoch": 2.6289739196649533, "grad_norm": 31.627735137939453, "learning_rate": 3.8447297975571285e-07, "loss": 0.5249, "step": 3454 }, { "epoch": 2.6297353893013518, "grad_norm": 29.246253967285156, "learning_rate": 3.829224664793574e-07, "loss": 0.6706, "step": 3455 }, { "epoch": 2.6304968589377498, "grad_norm": 30.79596519470215, "learning_rate": 3.81374961503575e-07, "loss": 0.6754, "step": 3456 }, { "epoch": 2.631258328574148, "grad_norm": 29.07205581665039, "learning_rate": 3.798304658366569e-07, "loss": 0.5742, "step": 3457 }, { "epoch": 2.632019798210546, "grad_norm": 31.699913024902344, "learning_rate": 3.782889804849316e-07, "loss": 0.5708, "step": 3458 }, { "epoch": 2.6327812678469447, "grad_norm": 39.25680923461914, "learning_rate": 3.767505064527688e-07, "loss": 0.6241, "step": 3459 }, { "epoch": 2.633542737483343, "grad_norm": 35.22846221923828, "learning_rate": 3.752150447425773e-07, "loss": 0.6787, "step": 3460 }, { "epoch": 2.634304207119741, "grad_norm": 31.9346981048584, "learning_rate": 3.736825963548013e-07, "loss": 0.5868, "step": 3461 }, { "epoch": 2.635065676756139, "grad_norm": 40.62056350708008, "learning_rate": 3.721531622879215e-07, "loss": 0.9302, "step": 3462 }, { "epoch": 2.6358271463925376, "grad_norm": 39.009700775146484, "learning_rate": 3.7062674353845585e-07, "loss": 0.7342, "step": 3463 }, { "epoch": 2.636588616028936, "grad_norm": 39.41543960571289, "learning_rate": 3.6910334110095594e-07, "loss": 0.6314, "step": 3464 }, { "epoch": 2.637350085665334, "grad_norm": 29.4053955078125, "learning_rate": 3.675829559680111e-07, "loss": 0.4331, "step": 3465 }, { "epoch": 2.6381115553017325, "grad_norm": 45.61473083496094, "learning_rate": 3.6606558913024003e-07, "loss": 0.7405, "step": 3466 }, { "epoch": 2.6388730249381305, "grad_norm": 31.86538314819336, "learning_rate": 3.6455124157629805e-07, "loss": 0.7126, "step": 3467 }, { "epoch": 2.639634494574529, "grad_norm": 39.75384521484375, "learning_rate": 3.630399142928742e-07, "loss": 0.6556, "step": 3468 }, { "epoch": 2.640395964210927, "grad_norm": 33.468624114990234, "learning_rate": 3.6153160826468694e-07, "loss": 0.6196, "step": 3469 }, { "epoch": 2.6411574338473254, "grad_norm": 35.19673538208008, "learning_rate": 3.6002632447448725e-07, "loss": 0.7563, "step": 3470 }, { "epoch": 2.641918903483724, "grad_norm": 31.554670333862305, "learning_rate": 3.5852406390305796e-07, "loss": 0.5347, "step": 3471 }, { "epoch": 2.642680373120122, "grad_norm": 150.344482421875, "learning_rate": 3.570248275292093e-07, "loss": 0.9876, "step": 3472 }, { "epoch": 2.64344184275652, "grad_norm": 37.18130111694336, "learning_rate": 3.555286163297855e-07, "loss": 0.5955, "step": 3473 }, { "epoch": 2.6442033123929183, "grad_norm": 35.696388244628906, "learning_rate": 3.5403543127965514e-07, "loss": 0.6277, "step": 3474 }, { "epoch": 2.6449647820293167, "grad_norm": 30.202611923217773, "learning_rate": 3.5254527335171764e-07, "loss": 0.6321, "step": 3475 }, { "epoch": 2.6457262516657147, "grad_norm": 38.498050689697266, "learning_rate": 3.5105814351689916e-07, "loss": 0.6903, "step": 3476 }, { "epoch": 2.646487721302113, "grad_norm": 38.22404098510742, "learning_rate": 3.495740427441541e-07, "loss": 0.7351, "step": 3477 }, { "epoch": 2.647249190938511, "grad_norm": 31.262937545776367, "learning_rate": 3.4809297200046166e-07, "loss": 0.5315, "step": 3478 }, { "epoch": 2.6480106605749096, "grad_norm": 35.6839599609375, "learning_rate": 3.4661493225082774e-07, "loss": 0.7766, "step": 3479 }, { "epoch": 2.6487721302113076, "grad_norm": 41.87835693359375, "learning_rate": 3.451399244582837e-07, "loss": 0.7617, "step": 3480 }, { "epoch": 2.649533599847706, "grad_norm": 37.20590591430664, "learning_rate": 3.436679495838835e-07, "loss": 0.5824, "step": 3481 }, { "epoch": 2.6502950694841045, "grad_norm": 39.00837326049805, "learning_rate": 3.4219900858670773e-07, "loss": 0.7096, "step": 3482 }, { "epoch": 2.6510565391205025, "grad_norm": 38.30519104003906, "learning_rate": 3.407331024238575e-07, "loss": 0.8263, "step": 3483 }, { "epoch": 2.6518180087569005, "grad_norm": 32.83727264404297, "learning_rate": 3.3927023205045875e-07, "loss": 0.6547, "step": 3484 }, { "epoch": 2.652579478393299, "grad_norm": 41.883567810058594, "learning_rate": 3.3781039841965733e-07, "loss": 0.6366, "step": 3485 }, { "epoch": 2.6533409480296974, "grad_norm": 27.319133758544922, "learning_rate": 3.363536024826236e-07, "loss": 0.5898, "step": 3486 }, { "epoch": 2.6541024176660954, "grad_norm": 30.66473960876465, "learning_rate": 3.3489984518854544e-07, "loss": 0.5786, "step": 3487 }, { "epoch": 2.654863887302494, "grad_norm": 32.9712028503418, "learning_rate": 3.334491274846324e-07, "loss": 0.7021, "step": 3488 }, { "epoch": 2.655625356938892, "grad_norm": 33.921592712402344, "learning_rate": 3.320014503161134e-07, "loss": 0.6292, "step": 3489 }, { "epoch": 2.6563868265752903, "grad_norm": 50.365970611572266, "learning_rate": 3.3055681462623613e-07, "loss": 0.8272, "step": 3490 }, { "epoch": 2.6571482962116884, "grad_norm": 50.954750061035156, "learning_rate": 3.2911522135626703e-07, "loss": 0.7422, "step": 3491 }, { "epoch": 2.657909765848087, "grad_norm": 42.99102020263672, "learning_rate": 3.2767667144548933e-07, "loss": 0.7505, "step": 3492 }, { "epoch": 2.6586712354844853, "grad_norm": 31.083498001098633, "learning_rate": 3.2624116583120315e-07, "loss": 0.6041, "step": 3493 }, { "epoch": 2.6594327051208833, "grad_norm": 33.25131607055664, "learning_rate": 3.2480870544872753e-07, "loss": 0.7391, "step": 3494 }, { "epoch": 2.6601941747572817, "grad_norm": 36.384376525878906, "learning_rate": 3.233792912313943e-07, "loss": 0.78, "step": 3495 }, { "epoch": 2.6609556443936797, "grad_norm": 38.22390365600586, "learning_rate": 3.2195292411055343e-07, "loss": 0.6149, "step": 3496 }, { "epoch": 2.661717114030078, "grad_norm": 33.089534759521484, "learning_rate": 3.2052960501556483e-07, "loss": 0.6382, "step": 3497 }, { "epoch": 2.662478583666476, "grad_norm": 44.604244232177734, "learning_rate": 3.1910933487380837e-07, "loss": 0.7966, "step": 3498 }, { "epoch": 2.6632400533028746, "grad_norm": 38.9262580871582, "learning_rate": 3.176921146106737e-07, "loss": 0.8145, "step": 3499 }, { "epoch": 2.664001522939273, "grad_norm": 38.19557189941406, "learning_rate": 3.1627794514956345e-07, "loss": 0.6644, "step": 3500 }, { "epoch": 2.664762992575671, "grad_norm": 39.922576904296875, "learning_rate": 3.148668274118938e-07, "loss": 0.7867, "step": 3501 }, { "epoch": 2.665524462212069, "grad_norm": 31.139284133911133, "learning_rate": 3.134587623170909e-07, "loss": 0.511, "step": 3502 }, { "epoch": 2.6662859318484675, "grad_norm": 30.733224868774414, "learning_rate": 3.120537507825949e-07, "loss": 0.7798, "step": 3503 }, { "epoch": 2.667047401484866, "grad_norm": 35.053619384765625, "learning_rate": 3.1065179372385343e-07, "loss": 0.7936, "step": 3504 }, { "epoch": 2.667808871121264, "grad_norm": 35.41645431518555, "learning_rate": 3.092528920543231e-07, "loss": 0.583, "step": 3505 }, { "epoch": 2.6685703407576624, "grad_norm": 44.019466400146484, "learning_rate": 3.0785704668547357e-07, "loss": 0.8197, "step": 3506 }, { "epoch": 2.6693318103940604, "grad_norm": 31.88688850402832, "learning_rate": 3.0646425852678074e-07, "loss": 0.6979, "step": 3507 }, { "epoch": 2.670093280030459, "grad_norm": 32.495731353759766, "learning_rate": 3.050745284857287e-07, "loss": 0.6334, "step": 3508 }, { "epoch": 2.670854749666857, "grad_norm": 33.78410339355469, "learning_rate": 3.0368785746780925e-07, "loss": 0.5368, "step": 3509 }, { "epoch": 2.6716162193032553, "grad_norm": 35.58042526245117, "learning_rate": 3.023042463765202e-07, "loss": 0.6375, "step": 3510 }, { "epoch": 2.6723776889396538, "grad_norm": 37.20764923095703, "learning_rate": 3.009236961133688e-07, "loss": 0.6795, "step": 3511 }, { "epoch": 2.6731391585760518, "grad_norm": 36.197566986083984, "learning_rate": 2.995462075778638e-07, "loss": 0.6648, "step": 3512 }, { "epoch": 2.67390062821245, "grad_norm": 30.067407608032227, "learning_rate": 2.981717816675206e-07, "loss": 0.5632, "step": 3513 }, { "epoch": 2.6746620978488482, "grad_norm": 30.508865356445312, "learning_rate": 2.968004192778595e-07, "loss": 0.5474, "step": 3514 }, { "epoch": 2.6754235674852467, "grad_norm": 39.49905776977539, "learning_rate": 2.9543212130240586e-07, "loss": 0.7179, "step": 3515 }, { "epoch": 2.6761850371216447, "grad_norm": 34.932559967041016, "learning_rate": 2.940668886326864e-07, "loss": 0.584, "step": 3516 }, { "epoch": 2.676946506758043, "grad_norm": 32.528385162353516, "learning_rate": 2.927047221582313e-07, "loss": 0.7726, "step": 3517 }, { "epoch": 2.677707976394441, "grad_norm": 35.287174224853516, "learning_rate": 2.913456227665734e-07, "loss": 0.6893, "step": 3518 }, { "epoch": 2.6784694460308396, "grad_norm": 37.9149169921875, "learning_rate": 2.8998959134324657e-07, "loss": 0.7816, "step": 3519 }, { "epoch": 2.6792309156672376, "grad_norm": 40.855342864990234, "learning_rate": 2.886366287717862e-07, "loss": 0.6403, "step": 3520 }, { "epoch": 2.679992385303636, "grad_norm": 35.35504913330078, "learning_rate": 2.8728673593372723e-07, "loss": 0.5108, "step": 3521 }, { "epoch": 2.6807538549400345, "grad_norm": 49.678382873535156, "learning_rate": 2.859399137086044e-07, "loss": 0.9165, "step": 3522 }, { "epoch": 2.6815153245764325, "grad_norm": 29.071609497070312, "learning_rate": 2.8459616297395464e-07, "loss": 0.6305, "step": 3523 }, { "epoch": 2.6822767942128305, "grad_norm": 41.9346809387207, "learning_rate": 2.832554846053104e-07, "loss": 0.8068, "step": 3524 }, { "epoch": 2.683038263849229, "grad_norm": 35.28245544433594, "learning_rate": 2.819178794762034e-07, "loss": 0.6542, "step": 3525 }, { "epoch": 2.6837997334856274, "grad_norm": 28.853044509887695, "learning_rate": 2.8058334845816214e-07, "loss": 0.6396, "step": 3526 }, { "epoch": 2.6845612031220254, "grad_norm": 27.49321174621582, "learning_rate": 2.792518924207144e-07, "loss": 0.5056, "step": 3527 }, { "epoch": 2.685322672758424, "grad_norm": 29.88329315185547, "learning_rate": 2.779235122313817e-07, "loss": 0.603, "step": 3528 }, { "epoch": 2.686084142394822, "grad_norm": 39.654563903808594, "learning_rate": 2.7659820875568345e-07, "loss": 0.806, "step": 3529 }, { "epoch": 2.6868456120312203, "grad_norm": 30.460655212402344, "learning_rate": 2.7527598285713387e-07, "loss": 0.6331, "step": 3530 }, { "epoch": 2.6876070816676183, "grad_norm": 28.07314682006836, "learning_rate": 2.7395683539723993e-07, "loss": 0.5977, "step": 3531 }, { "epoch": 2.6883685513040168, "grad_norm": 32.25593566894531, "learning_rate": 2.7264076723550694e-07, "loss": 0.5636, "step": 3532 }, { "epoch": 2.689130020940415, "grad_norm": 31.922256469726562, "learning_rate": 2.713277792294311e-07, "loss": 0.7314, "step": 3533 }, { "epoch": 2.689891490576813, "grad_norm": 44.6668586730957, "learning_rate": 2.700178722345015e-07, "loss": 0.8454, "step": 3534 }, { "epoch": 2.6906529602132117, "grad_norm": 33.710777282714844, "learning_rate": 2.6871104710420037e-07, "loss": 0.66, "step": 3535 }, { "epoch": 2.6914144298496097, "grad_norm": 28.016494750976562, "learning_rate": 2.674073046900028e-07, "loss": 0.4529, "step": 3536 }, { "epoch": 2.692175899486008, "grad_norm": 43.784725189208984, "learning_rate": 2.6610664584137413e-07, "loss": 0.7916, "step": 3537 }, { "epoch": 2.692937369122406, "grad_norm": 38.62266540527344, "learning_rate": 2.648090714057705e-07, "loss": 0.6615, "step": 3538 }, { "epoch": 2.6936988387588046, "grad_norm": 34.8658561706543, "learning_rate": 2.635145822286389e-07, "loss": 0.5677, "step": 3539 }, { "epoch": 2.694460308395203, "grad_norm": 36.61354446411133, "learning_rate": 2.622231791534158e-07, "loss": 0.6602, "step": 3540 }, { "epoch": 2.695221778031601, "grad_norm": 31.112960815429688, "learning_rate": 2.609348630215286e-07, "loss": 0.5554, "step": 3541 }, { "epoch": 2.695983247667999, "grad_norm": 37.42208480834961, "learning_rate": 2.5964963467239034e-07, "loss": 0.7106, "step": 3542 }, { "epoch": 2.6967447173043975, "grad_norm": 37.736148834228516, "learning_rate": 2.583674949434045e-07, "loss": 0.6659, "step": 3543 }, { "epoch": 2.697506186940796, "grad_norm": 40.0212516784668, "learning_rate": 2.570884446699612e-07, "loss": 0.7413, "step": 3544 }, { "epoch": 2.698267656577194, "grad_norm": 33.9966926574707, "learning_rate": 2.558124846854371e-07, "loss": 0.663, "step": 3545 }, { "epoch": 2.6990291262135924, "grad_norm": 45.15996170043945, "learning_rate": 2.5453961582119715e-07, "loss": 0.598, "step": 3546 }, { "epoch": 2.6997905958499904, "grad_norm": 35.7940788269043, "learning_rate": 2.532698389065902e-07, "loss": 0.6968, "step": 3547 }, { "epoch": 2.700552065486389, "grad_norm": 37.959144592285156, "learning_rate": 2.520031547689505e-07, "loss": 0.7327, "step": 3548 }, { "epoch": 2.701313535122787, "grad_norm": 42.76571273803711, "learning_rate": 2.507395642336008e-07, "loss": 0.6189, "step": 3549 }, { "epoch": 2.7020750047591853, "grad_norm": 38.12059020996094, "learning_rate": 2.4947906812384404e-07, "loss": 0.7697, "step": 3550 }, { "epoch": 2.7028364743955837, "grad_norm": 42.39665222167969, "learning_rate": 2.482216672609677e-07, "loss": 0.8878, "step": 3551 }, { "epoch": 2.7035979440319817, "grad_norm": 31.990509033203125, "learning_rate": 2.469673624642438e-07, "loss": 0.6218, "step": 3552 }, { "epoch": 2.7043594136683797, "grad_norm": 30.26728057861328, "learning_rate": 2.457161545509268e-07, "loss": 0.643, "step": 3553 }, { "epoch": 2.705120883304778, "grad_norm": 38.14008712768555, "learning_rate": 2.4446804433625235e-07, "loss": 0.7505, "step": 3554 }, { "epoch": 2.7058823529411766, "grad_norm": 47.42528533935547, "learning_rate": 2.4322303263343947e-07, "loss": 0.8773, "step": 3555 }, { "epoch": 2.7066438225775746, "grad_norm": 33.20399475097656, "learning_rate": 2.4198112025368646e-07, "loss": 0.6999, "step": 3556 }, { "epoch": 2.707405292213973, "grad_norm": 38.39046096801758, "learning_rate": 2.4074230800617273e-07, "loss": 0.86, "step": 3557 }, { "epoch": 2.708166761850371, "grad_norm": 28.45980453491211, "learning_rate": 2.3950659669806033e-07, "loss": 0.5588, "step": 3558 }, { "epoch": 2.7089282314867695, "grad_norm": 38.409732818603516, "learning_rate": 2.3827398713448635e-07, "loss": 0.683, "step": 3559 }, { "epoch": 2.7096897011231675, "grad_norm": 32.38550567626953, "learning_rate": 2.370444801185695e-07, "loss": 0.5327, "step": 3560 }, { "epoch": 2.710451170759566, "grad_norm": 42.81605911254883, "learning_rate": 2.3581807645140785e-07, "loss": 0.8608, "step": 3561 }, { "epoch": 2.7112126403959644, "grad_norm": 33.16263961791992, "learning_rate": 2.3459477693207556e-07, "loss": 0.5887, "step": 3562 }, { "epoch": 2.7119741100323624, "grad_norm": 40.236488342285156, "learning_rate": 2.33374582357625e-07, "loss": 0.7936, "step": 3563 }, { "epoch": 2.7127355796687604, "grad_norm": 31.314218521118164, "learning_rate": 2.321574935230858e-07, "loss": 0.6352, "step": 3564 }, { "epoch": 2.713497049305159, "grad_norm": 44.145206451416016, "learning_rate": 2.3094351122146307e-07, "loss": 0.7156, "step": 3565 }, { "epoch": 2.7142585189415573, "grad_norm": 34.7823371887207, "learning_rate": 2.2973263624373965e-07, "loss": 0.6705, "step": 3566 }, { "epoch": 2.7150199885779553, "grad_norm": 39.29833221435547, "learning_rate": 2.2852486937887163e-07, "loss": 0.7412, "step": 3567 }, { "epoch": 2.715781458214354, "grad_norm": 35.539302825927734, "learning_rate": 2.2732021141379124e-07, "loss": 0.6821, "step": 3568 }, { "epoch": 2.716542927850752, "grad_norm": 34.4547233581543, "learning_rate": 2.2611866313340391e-07, "loss": 0.7134, "step": 3569 }, { "epoch": 2.7173043974871502, "grad_norm": 45.34823226928711, "learning_rate": 2.2492022532059122e-07, "loss": 0.7329, "step": 3570 }, { "epoch": 2.7180658671235483, "grad_norm": 33.655860900878906, "learning_rate": 2.237248987562063e-07, "loss": 0.6694, "step": 3571 }, { "epoch": 2.7188273367599467, "grad_norm": 38.89566421508789, "learning_rate": 2.2253268421907503e-07, "loss": 0.7674, "step": 3572 }, { "epoch": 2.719588806396345, "grad_norm": 29.599313735961914, "learning_rate": 2.2134358248599663e-07, "loss": 0.5496, "step": 3573 }, { "epoch": 2.720350276032743, "grad_norm": 33.48070526123047, "learning_rate": 2.2015759433174079e-07, "loss": 0.6482, "step": 3574 }, { "epoch": 2.7211117456691416, "grad_norm": 28.72620391845703, "learning_rate": 2.189747205290499e-07, "loss": 0.5613, "step": 3575 }, { "epoch": 2.7218732153055396, "grad_norm": 46.34340286254883, "learning_rate": 2.1779496184863634e-07, "loss": 0.9551, "step": 3576 }, { "epoch": 2.722634684941938, "grad_norm": 38.35898208618164, "learning_rate": 2.1661831905918363e-07, "loss": 0.6016, "step": 3577 }, { "epoch": 2.723396154578336, "grad_norm": 32.96267318725586, "learning_rate": 2.1544479292734344e-07, "loss": 0.6609, "step": 3578 }, { "epoch": 2.7241576242147345, "grad_norm": 39.011592864990234, "learning_rate": 2.142743842177386e-07, "loss": 0.7328, "step": 3579 }, { "epoch": 2.724919093851133, "grad_norm": 33.40355682373047, "learning_rate": 2.1310709369296022e-07, "loss": 0.6519, "step": 3580 }, { "epoch": 2.725680563487531, "grad_norm": 34.59334182739258, "learning_rate": 2.1194292211356715e-07, "loss": 0.651, "step": 3581 }, { "epoch": 2.726442033123929, "grad_norm": 42.002655029296875, "learning_rate": 2.1078187023808595e-07, "loss": 0.7626, "step": 3582 }, { "epoch": 2.7272035027603274, "grad_norm": 42.294593811035156, "learning_rate": 2.0962393882301148e-07, "loss": 0.8532, "step": 3583 }, { "epoch": 2.727964972396726, "grad_norm": 33.331233978271484, "learning_rate": 2.0846912862280522e-07, "loss": 0.5851, "step": 3584 }, { "epoch": 2.728726442033124, "grad_norm": 41.88325881958008, "learning_rate": 2.0731744038989364e-07, "loss": 0.8611, "step": 3585 }, { "epoch": 2.7294879116695223, "grad_norm": 40.594425201416016, "learning_rate": 2.0616887487467096e-07, "loss": 0.742, "step": 3586 }, { "epoch": 2.7302493813059203, "grad_norm": 37.78084182739258, "learning_rate": 2.0502343282549576e-07, "loss": 0.5889, "step": 3587 }, { "epoch": 2.7310108509423188, "grad_norm": 35.8529167175293, "learning_rate": 2.0388111498869157e-07, "loss": 0.7138, "step": 3588 }, { "epoch": 2.7317723205787168, "grad_norm": 29.707866668701172, "learning_rate": 2.0274192210854694e-07, "loss": 0.5175, "step": 3589 }, { "epoch": 2.732533790215115, "grad_norm": 43.78036880493164, "learning_rate": 2.0160585492731254e-07, "loss": 0.7013, "step": 3590 }, { "epoch": 2.7332952598515137, "grad_norm": 30.173080444335938, "learning_rate": 2.0047291418520466e-07, "loss": 0.6849, "step": 3591 }, { "epoch": 2.7340567294879117, "grad_norm": 40.91156005859375, "learning_rate": 1.9934310062040108e-07, "loss": 0.7347, "step": 3592 }, { "epoch": 2.7348181991243097, "grad_norm": 34.31969451904297, "learning_rate": 1.98216414969043e-07, "loss": 0.7648, "step": 3593 }, { "epoch": 2.735579668760708, "grad_norm": 33.60858154296875, "learning_rate": 1.970928579652326e-07, "loss": 0.7227, "step": 3594 }, { "epoch": 2.7363411383971066, "grad_norm": 30.68975830078125, "learning_rate": 1.959724303410343e-07, "loss": 0.4282, "step": 3595 }, { "epoch": 2.7371026080335046, "grad_norm": 30.861906051635742, "learning_rate": 1.9485513282647361e-07, "loss": 0.6189, "step": 3596 }, { "epoch": 2.737864077669903, "grad_norm": 33.06851577758789, "learning_rate": 1.9374096614953652e-07, "loss": 0.6854, "step": 3597 }, { "epoch": 2.738625547306301, "grad_norm": 43.72724914550781, "learning_rate": 1.9262993103616734e-07, "loss": 0.8647, "step": 3598 }, { "epoch": 2.7393870169426995, "grad_norm": 38.38490295410156, "learning_rate": 1.915220282102731e-07, "loss": 0.7559, "step": 3599 }, { "epoch": 2.7401484865790975, "grad_norm": 54.58589553833008, "learning_rate": 1.9041725839371805e-07, "loss": 0.902, "step": 3600 }, { "epoch": 2.740909956215496, "grad_norm": 40.822662353515625, "learning_rate": 1.893156223063253e-07, "loss": 0.6479, "step": 3601 }, { "epoch": 2.7416714258518944, "grad_norm": 43.5707893371582, "learning_rate": 1.8821712066587617e-07, "loss": 0.9983, "step": 3602 }, { "epoch": 2.7424328954882924, "grad_norm": 32.73143768310547, "learning_rate": 1.871217541881093e-07, "loss": 0.6216, "step": 3603 }, { "epoch": 2.7431943651246904, "grad_norm": 34.381038665771484, "learning_rate": 1.8602952358672212e-07, "loss": 0.7752, "step": 3604 }, { "epoch": 2.743955834761089, "grad_norm": 35.17359924316406, "learning_rate": 1.8494042957336754e-07, "loss": 0.6721, "step": 3605 }, { "epoch": 2.7447173043974873, "grad_norm": 42.144771575927734, "learning_rate": 1.8385447285765468e-07, "loss": 0.6494, "step": 3606 }, { "epoch": 2.7454787740338853, "grad_norm": 44.28874588012695, "learning_rate": 1.8277165414714858e-07, "loss": 0.9447, "step": 3607 }, { "epoch": 2.7462402436702837, "grad_norm": 34.879878997802734, "learning_rate": 1.8169197414737105e-07, "loss": 0.7277, "step": 3608 }, { "epoch": 2.7470017133066817, "grad_norm": 37.577754974365234, "learning_rate": 1.806154335617971e-07, "loss": 0.7279, "step": 3609 }, { "epoch": 2.74776318294308, "grad_norm": 36.11455535888672, "learning_rate": 1.7954203309185736e-07, "loss": 0.7481, "step": 3610 }, { "epoch": 2.748524652579478, "grad_norm": 31.323657989501953, "learning_rate": 1.7847177343693568e-07, "loss": 0.5463, "step": 3611 }, { "epoch": 2.7492861222158766, "grad_norm": 25.571853637695312, "learning_rate": 1.7740465529436924e-07, "loss": 0.6026, "step": 3612 }, { "epoch": 2.750047591852275, "grad_norm": 42.7362174987793, "learning_rate": 1.7634067935945133e-07, "loss": 0.9733, "step": 3613 }, { "epoch": 2.750809061488673, "grad_norm": 41.02425765991211, "learning_rate": 1.752798463254235e-07, "loss": 0.7146, "step": 3614 }, { "epoch": 2.7515705311250716, "grad_norm": 31.332490921020508, "learning_rate": 1.7422215688348177e-07, "loss": 0.7065, "step": 3615 }, { "epoch": 2.7523320007614696, "grad_norm": 30.479360580444336, "learning_rate": 1.7316761172277319e-07, "loss": 0.4783, "step": 3616 }, { "epoch": 2.753093470397868, "grad_norm": 30.810632705688477, "learning_rate": 1.721162115303987e-07, "loss": 0.5426, "step": 3617 }, { "epoch": 2.753854940034266, "grad_norm": 23.39420509338379, "learning_rate": 1.7106795699140588e-07, "loss": 0.5192, "step": 3618 }, { "epoch": 2.7546164096706645, "grad_norm": 31.11192512512207, "learning_rate": 1.7002284878879617e-07, "loss": 0.662, "step": 3619 }, { "epoch": 2.755377879307063, "grad_norm": 34.58052062988281, "learning_rate": 1.6898088760351983e-07, "loss": 0.7205, "step": 3620 }, { "epoch": 2.756139348943461, "grad_norm": 38.402828216552734, "learning_rate": 1.6794207411447548e-07, "loss": 0.7301, "step": 3621 }, { "epoch": 2.756900818579859, "grad_norm": 32.64921951293945, "learning_rate": 1.669064089985123e-07, "loss": 0.7054, "step": 3622 }, { "epoch": 2.7576622882162574, "grad_norm": 39.262813568115234, "learning_rate": 1.658738929304282e-07, "loss": 0.726, "step": 3623 }, { "epoch": 2.758423757852656, "grad_norm": 26.34572982788086, "learning_rate": 1.648445265829679e-07, "loss": 0.5428, "step": 3624 }, { "epoch": 2.759185227489054, "grad_norm": 25.74113655090332, "learning_rate": 1.638183106268254e-07, "loss": 0.675, "step": 3625 }, { "epoch": 2.7599466971254523, "grad_norm": 31.403913497924805, "learning_rate": 1.6279524573064142e-07, "loss": 0.6615, "step": 3626 }, { "epoch": 2.7607081667618503, "grad_norm": 37.42214584350586, "learning_rate": 1.6177533256100386e-07, "loss": 0.6695, "step": 3627 }, { "epoch": 2.7614696363982487, "grad_norm": 35.03535842895508, "learning_rate": 1.6075857178244613e-07, "loss": 0.6955, "step": 3628 }, { "epoch": 2.7622311060346467, "grad_norm": 27.508697509765625, "learning_rate": 1.5974496405744888e-07, "loss": 0.4932, "step": 3629 }, { "epoch": 2.762992575671045, "grad_norm": 33.885948181152344, "learning_rate": 1.587345100464377e-07, "loss": 0.63, "step": 3630 }, { "epoch": 2.7637540453074436, "grad_norm": 26.789216995239258, "learning_rate": 1.5772721040778317e-07, "loss": 0.6555, "step": 3631 }, { "epoch": 2.7645155149438416, "grad_norm": 37.04738998413086, "learning_rate": 1.5672306579780084e-07, "loss": 0.7001, "step": 3632 }, { "epoch": 2.7652769845802396, "grad_norm": 44.23310852050781, "learning_rate": 1.557220768707507e-07, "loss": 0.8529, "step": 3633 }, { "epoch": 2.766038454216638, "grad_norm": 42.33627700805664, "learning_rate": 1.5472424427883715e-07, "loss": 0.7125, "step": 3634 }, { "epoch": 2.7667999238530365, "grad_norm": 44.1581916809082, "learning_rate": 1.5372956867220678e-07, "loss": 0.7352, "step": 3635 }, { "epoch": 2.7675613934894345, "grad_norm": 43.5903205871582, "learning_rate": 1.5273805069895065e-07, "loss": 0.8347, "step": 3636 }, { "epoch": 2.768322863125833, "grad_norm": 43.04545974731445, "learning_rate": 1.5174969100510084e-07, "loss": 0.7801, "step": 3637 }, { "epoch": 2.769084332762231, "grad_norm": 32.42032241821289, "learning_rate": 1.507644902346328e-07, "loss": 0.7264, "step": 3638 }, { "epoch": 2.7698458023986294, "grad_norm": 36.454925537109375, "learning_rate": 1.497824490294636e-07, "loss": 0.8303, "step": 3639 }, { "epoch": 2.7706072720350274, "grad_norm": 30.78838539123535, "learning_rate": 1.488035680294514e-07, "loss": 0.507, "step": 3640 }, { "epoch": 2.771368741671426, "grad_norm": 33.48661422729492, "learning_rate": 1.478278478723949e-07, "loss": 0.7411, "step": 3641 }, { "epoch": 2.7721302113078243, "grad_norm": 41.15423583984375, "learning_rate": 1.4685528919403446e-07, "loss": 0.622, "step": 3642 }, { "epoch": 2.7728916809442223, "grad_norm": 36.249366760253906, "learning_rate": 1.4588589262804986e-07, "loss": 0.5892, "step": 3643 }, { "epoch": 2.7736531505806203, "grad_norm": 28.01523780822754, "learning_rate": 1.4491965880606086e-07, "loss": 0.7187, "step": 3644 }, { "epoch": 2.774414620217019, "grad_norm": 28.708953857421875, "learning_rate": 1.439565883576244e-07, "loss": 0.5966, "step": 3645 }, { "epoch": 2.7751760898534172, "grad_norm": 33.08787536621094, "learning_rate": 1.429966819102402e-07, "loss": 0.616, "step": 3646 }, { "epoch": 2.7759375594898152, "grad_norm": 38.922119140625, "learning_rate": 1.4203994008934407e-07, "loss": 0.8565, "step": 3647 }, { "epoch": 2.7766990291262137, "grad_norm": 30.668699264526367, "learning_rate": 1.4108636351830962e-07, "loss": 0.5661, "step": 3648 }, { "epoch": 2.7774604987626117, "grad_norm": 30.796104431152344, "learning_rate": 1.4013595281844872e-07, "loss": 0.6419, "step": 3649 }, { "epoch": 2.77822196839901, "grad_norm": 30.39305877685547, "learning_rate": 1.3918870860901046e-07, "loss": 0.4922, "step": 3650 }, { "epoch": 2.778983438035408, "grad_norm": 37.66404342651367, "learning_rate": 1.3824463150718114e-07, "loss": 0.7657, "step": 3651 }, { "epoch": 2.7797449076718066, "grad_norm": 43.04560470581055, "learning_rate": 1.3730372212808374e-07, "loss": 0.6948, "step": 3652 }, { "epoch": 2.780506377308205, "grad_norm": 41.97024917602539, "learning_rate": 1.3636598108477506e-07, "loss": 0.8338, "step": 3653 }, { "epoch": 2.781267846944603, "grad_norm": 36.9130744934082, "learning_rate": 1.3543140898825024e-07, "loss": 0.5565, "step": 3654 }, { "epoch": 2.7820293165810015, "grad_norm": 30.372880935668945, "learning_rate": 1.3450000644743831e-07, "loss": 0.5575, "step": 3655 }, { "epoch": 2.7827907862173995, "grad_norm": 35.18635177612305, "learning_rate": 1.335717740692033e-07, "loss": 0.6885, "step": 3656 }, { "epoch": 2.783552255853798, "grad_norm": 31.250743865966797, "learning_rate": 1.3264671245834416e-07, "loss": 0.6199, "step": 3657 }, { "epoch": 2.784313725490196, "grad_norm": 40.596046447753906, "learning_rate": 1.3172482221759264e-07, "loss": 0.6757, "step": 3658 }, { "epoch": 2.7850751951265944, "grad_norm": 40.5785026550293, "learning_rate": 1.3080610394761717e-07, "loss": 0.7238, "step": 3659 }, { "epoch": 2.7858366647629924, "grad_norm": 42.478843688964844, "learning_rate": 1.2989055824701502e-07, "loss": 0.7429, "step": 3660 }, { "epoch": 2.786598134399391, "grad_norm": 39.95835876464844, "learning_rate": 1.2897818571231957e-07, "loss": 0.7592, "step": 3661 }, { "epoch": 2.787359604035789, "grad_norm": 30.3173770904541, "learning_rate": 1.2806898693799474e-07, "loss": 0.522, "step": 3662 }, { "epoch": 2.7881210736721873, "grad_norm": 32.92494583129883, "learning_rate": 1.2716296251644e-07, "loss": 0.6268, "step": 3663 }, { "epoch": 2.7888825433085858, "grad_norm": 38.015960693359375, "learning_rate": 1.2626011303798202e-07, "loss": 0.8905, "step": 3664 }, { "epoch": 2.7896440129449838, "grad_norm": 34.898887634277344, "learning_rate": 1.253604390908819e-07, "loss": 0.4904, "step": 3665 }, { "epoch": 2.790405482581382, "grad_norm": 31.011995315551758, "learning_rate": 1.2446394126133078e-07, "loss": 0.7325, "step": 3666 }, { "epoch": 2.79116695221778, "grad_norm": 44.569149017333984, "learning_rate": 1.2357062013344968e-07, "loss": 0.7484, "step": 3667 }, { "epoch": 2.7919284218541787, "grad_norm": 31.56942367553711, "learning_rate": 1.226804762892908e-07, "loss": 0.6672, "step": 3668 }, { "epoch": 2.7926898914905767, "grad_norm": 34.807586669921875, "learning_rate": 1.217935103088358e-07, "loss": 0.6472, "step": 3669 }, { "epoch": 2.793451361126975, "grad_norm": 37.42338943481445, "learning_rate": 1.2090972276999513e-07, "loss": 0.7673, "step": 3670 }, { "epoch": 2.7942128307633736, "grad_norm": 52.045684814453125, "learning_rate": 1.2002911424860985e-07, "loss": 0.84, "step": 3671 }, { "epoch": 2.7949743003997716, "grad_norm": 36.23496627807617, "learning_rate": 1.1915168531844823e-07, "loss": 0.6209, "step": 3672 }, { "epoch": 2.7957357700361696, "grad_norm": 39.331565856933594, "learning_rate": 1.1827743655120794e-07, "loss": 0.6987, "step": 3673 }, { "epoch": 2.796497239672568, "grad_norm": 35.434181213378906, "learning_rate": 1.1740636851651333e-07, "loss": 0.7311, "step": 3674 }, { "epoch": 2.7972587093089665, "grad_norm": 39.36561965942383, "learning_rate": 1.1653848178191763e-07, "loss": 0.8459, "step": 3675 }, { "epoch": 2.7980201789453645, "grad_norm": 42.74898147583008, "learning_rate": 1.1567377691290072e-07, "loss": 0.7566, "step": 3676 }, { "epoch": 2.798781648581763, "grad_norm": 31.13348388671875, "learning_rate": 1.1481225447286803e-07, "loss": 0.4804, "step": 3677 }, { "epoch": 2.799543118218161, "grad_norm": 36.430667877197266, "learning_rate": 1.1395391502315445e-07, "loss": 0.7796, "step": 3678 }, { "epoch": 2.8003045878545594, "grad_norm": 36.631004333496094, "learning_rate": 1.1309875912301704e-07, "loss": 0.7551, "step": 3679 }, { "epoch": 2.8010660574909574, "grad_norm": 36.21662139892578, "learning_rate": 1.1224678732964289e-07, "loss": 0.6659, "step": 3680 }, { "epoch": 2.801827527127356, "grad_norm": 27.77839469909668, "learning_rate": 1.1139800019814128e-07, "loss": 0.6801, "step": 3681 }, { "epoch": 2.8025889967637543, "grad_norm": 39.974327087402344, "learning_rate": 1.1055239828154762e-07, "loss": 0.7036, "step": 3682 }, { "epoch": 2.8033504664001523, "grad_norm": 39.37786102294922, "learning_rate": 1.0970998213082173e-07, "loss": 0.7469, "step": 3683 }, { "epoch": 2.8041119360365503, "grad_norm": 46.03701400756836, "learning_rate": 1.0887075229484789e-07, "loss": 0.7971, "step": 3684 }, { "epoch": 2.8048734056729487, "grad_norm": 30.099138259887695, "learning_rate": 1.0803470932043425e-07, "loss": 0.6927, "step": 3685 }, { "epoch": 2.805634875309347, "grad_norm": 34.72941970825195, "learning_rate": 1.0720185375231284e-07, "loss": 0.6556, "step": 3686 }, { "epoch": 2.806396344945745, "grad_norm": 40.67279052734375, "learning_rate": 1.0637218613313793e-07, "loss": 0.6538, "step": 3687 }, { "epoch": 2.8071578145821436, "grad_norm": 35.082035064697266, "learning_rate": 1.055457070034871e-07, "loss": 0.6581, "step": 3688 }, { "epoch": 2.8079192842185416, "grad_norm": 34.50794219970703, "learning_rate": 1.047224169018618e-07, "loss": 0.7243, "step": 3689 }, { "epoch": 2.80868075385494, "grad_norm": 43.9166374206543, "learning_rate": 1.0390231636468406e-07, "loss": 0.6786, "step": 3690 }, { "epoch": 2.809442223491338, "grad_norm": 33.39232635498047, "learning_rate": 1.0308540592629756e-07, "loss": 0.7364, "step": 3691 }, { "epoch": 2.8102036931277365, "grad_norm": 47.794769287109375, "learning_rate": 1.0227168611896876e-07, "loss": 0.7986, "step": 3692 }, { "epoch": 2.810965162764135, "grad_norm": 39.57087707519531, "learning_rate": 1.0146115747288409e-07, "loss": 0.6943, "step": 3693 }, { "epoch": 2.811726632400533, "grad_norm": 34.416595458984375, "learning_rate": 1.0065382051615114e-07, "loss": 0.7921, "step": 3694 }, { "epoch": 2.812488102036931, "grad_norm": 37.67659378051758, "learning_rate": 9.984967577479854e-08, "loss": 0.8022, "step": 3695 }, { "epoch": 2.8132495716733295, "grad_norm": 42.646263122558594, "learning_rate": 9.904872377277275e-08, "loss": 0.7815, "step": 3696 }, { "epoch": 2.814011041309728, "grad_norm": 38.440773010253906, "learning_rate": 9.825096503194409e-08, "loss": 0.6546, "step": 3697 }, { "epoch": 2.814772510946126, "grad_norm": 41.80242919921875, "learning_rate": 9.745640007209844e-08, "loss": 0.7553, "step": 3698 }, { "epoch": 2.8155339805825244, "grad_norm": 35.23366928100586, "learning_rate": 9.666502941094169e-08, "loss": 0.5881, "step": 3699 }, { "epoch": 2.8162954502189224, "grad_norm": 36.92134094238281, "learning_rate": 9.587685356409914e-08, "loss": 0.6313, "step": 3700 }, { "epoch": 2.817056919855321, "grad_norm": 43.630279541015625, "learning_rate": 9.509187304511502e-08, "loss": 0.8311, "step": 3701 }, { "epoch": 2.817818389491719, "grad_norm": 30.255191802978516, "learning_rate": 9.43100883654502e-08, "loss": 0.5692, "step": 3702 }, { "epoch": 2.8185798591281173, "grad_norm": 26.731603622436523, "learning_rate": 9.353150003448441e-08, "loss": 0.5962, "step": 3703 }, { "epoch": 2.8193413287645157, "grad_norm": 34.332984924316406, "learning_rate": 9.275610855951301e-08, "loss": 0.7872, "step": 3704 }, { "epoch": 2.8201027984009137, "grad_norm": 32.72539138793945, "learning_rate": 9.198391444575072e-08, "loss": 0.629, "step": 3705 }, { "epoch": 2.820864268037312, "grad_norm": 36.383113861083984, "learning_rate": 9.121491819632844e-08, "loss": 0.8126, "step": 3706 }, { "epoch": 2.82162573767371, "grad_norm": 29.912612915039062, "learning_rate": 9.044912031229147e-08, "loss": 0.5893, "step": 3707 }, { "epoch": 2.8223872073101086, "grad_norm": 43.911460876464844, "learning_rate": 8.968652129260347e-08, "loss": 0.6809, "step": 3708 }, { "epoch": 2.8231486769465066, "grad_norm": 41.07213592529297, "learning_rate": 8.892712163414252e-08, "loss": 0.6893, "step": 3709 }, { "epoch": 2.823910146582905, "grad_norm": 33.28770065307617, "learning_rate": 8.817092183170395e-08, "loss": 0.6444, "step": 3710 }, { "epoch": 2.8246716162193035, "grad_norm": 34.47660827636719, "learning_rate": 8.741792237799584e-08, "loss": 0.6877, "step": 3711 }, { "epoch": 2.8254330858557015, "grad_norm": 31.142662048339844, "learning_rate": 8.666812376364187e-08, "loss": 0.6381, "step": 3712 }, { "epoch": 2.8261945554920995, "grad_norm": 37.53310012817383, "learning_rate": 8.592152647718066e-08, "loss": 0.9058, "step": 3713 }, { "epoch": 2.826956025128498, "grad_norm": 40.25123596191406, "learning_rate": 8.517813100506644e-08, "loss": 0.7142, "step": 3714 }, { "epoch": 2.8277174947648964, "grad_norm": 33.80186080932617, "learning_rate": 8.443793783166287e-08, "loss": 0.7497, "step": 3715 }, { "epoch": 2.8284789644012944, "grad_norm": 43.681888580322266, "learning_rate": 8.370094743925139e-08, "loss": 0.9071, "step": 3716 }, { "epoch": 2.829240434037693, "grad_norm": 37.438804626464844, "learning_rate": 8.296716030802454e-08, "loss": 0.7432, "step": 3717 }, { "epoch": 2.830001903674091, "grad_norm": 34.2763671875, "learning_rate": 8.223657691608822e-08, "loss": 0.7006, "step": 3718 }, { "epoch": 2.8307633733104893, "grad_norm": 37.89769744873047, "learning_rate": 8.150919773946165e-08, "loss": 0.7898, "step": 3719 }, { "epoch": 2.8315248429468873, "grad_norm": 39.76044845581055, "learning_rate": 8.078502325207516e-08, "loss": 0.7265, "step": 3720 }, { "epoch": 2.832286312583286, "grad_norm": 40.89030075073242, "learning_rate": 8.006405392577132e-08, "loss": 0.6904, "step": 3721 }, { "epoch": 2.8330477822196842, "grad_norm": 34.278743743896484, "learning_rate": 7.934629023030493e-08, "loss": 0.6828, "step": 3722 }, { "epoch": 2.8338092518560822, "grad_norm": 33.07197189331055, "learning_rate": 7.863173263334078e-08, "loss": 0.6427, "step": 3723 }, { "epoch": 2.8345707214924802, "grad_norm": 55.87440490722656, "learning_rate": 7.7920381600457e-08, "loss": 0.6936, "step": 3724 }, { "epoch": 2.8353321911288787, "grad_norm": 27.274700164794922, "learning_rate": 7.721223759514007e-08, "loss": 0.5088, "step": 3725 }, { "epoch": 2.836093660765277, "grad_norm": 39.85972595214844, "learning_rate": 7.650730107878812e-08, "loss": 0.7226, "step": 3726 }, { "epoch": 2.836855130401675, "grad_norm": 37.053775787353516, "learning_rate": 7.580557251070985e-08, "loss": 0.9435, "step": 3727 }, { "epoch": 2.8376166000380736, "grad_norm": 44.19858932495117, "learning_rate": 7.510705234812232e-08, "loss": 0.9758, "step": 3728 }, { "epoch": 2.8383780696744716, "grad_norm": 34.89821243286133, "learning_rate": 7.441174104615368e-08, "loss": 0.5827, "step": 3729 }, { "epoch": 2.83913953931087, "grad_norm": 38.33170700073242, "learning_rate": 7.371963905783986e-08, "loss": 0.6014, "step": 3730 }, { "epoch": 2.839901008947268, "grad_norm": 28.10875129699707, "learning_rate": 7.30307468341268e-08, "loss": 0.5353, "step": 3731 }, { "epoch": 2.8406624785836665, "grad_norm": 35.89402770996094, "learning_rate": 7.234506482386883e-08, "loss": 0.6917, "step": 3732 }, { "epoch": 2.841423948220065, "grad_norm": 28.482637405395508, "learning_rate": 7.166259347382854e-08, "loss": 0.5436, "step": 3733 }, { "epoch": 2.842185417856463, "grad_norm": 33.73384094238281, "learning_rate": 7.098333322867579e-08, "loss": 0.8096, "step": 3734 }, { "epoch": 2.842946887492861, "grad_norm": 51.8864860534668, "learning_rate": 7.030728453099046e-08, "loss": 0.8937, "step": 3735 }, { "epoch": 2.8437083571292594, "grad_norm": 34.431365966796875, "learning_rate": 6.963444782125794e-08, "loss": 0.7507, "step": 3736 }, { "epoch": 2.844469826765658, "grad_norm": 41.20650100708008, "learning_rate": 6.896482353787148e-08, "loss": 0.8473, "step": 3737 }, { "epoch": 2.845231296402056, "grad_norm": 33.853275299072266, "learning_rate": 6.829841211712984e-08, "loss": 0.6342, "step": 3738 }, { "epoch": 2.8459927660384543, "grad_norm": 41.89827346801758, "learning_rate": 6.763521399324124e-08, "loss": 0.713, "step": 3739 }, { "epoch": 2.8467542356748523, "grad_norm": 32.10409164428711, "learning_rate": 6.697522959831837e-08, "loss": 0.669, "step": 3740 }, { "epoch": 2.8475157053112508, "grad_norm": 33.64268493652344, "learning_rate": 6.631845936238001e-08, "loss": 0.5775, "step": 3741 }, { "epoch": 2.8482771749476488, "grad_norm": 33.45774459838867, "learning_rate": 6.566490371335055e-08, "loss": 0.7358, "step": 3742 }, { "epoch": 2.849038644584047, "grad_norm": 30.134992599487305, "learning_rate": 6.5014563077061e-08, "loss": 0.6463, "step": 3743 }, { "epoch": 2.8498001142204457, "grad_norm": 41.09950256347656, "learning_rate": 6.436743787724741e-08, "loss": 0.8201, "step": 3744 }, { "epoch": 2.8505615838568437, "grad_norm": 33.73673629760742, "learning_rate": 6.372352853554975e-08, "loss": 0.7357, "step": 3745 }, { "epoch": 2.851323053493242, "grad_norm": 32.61793518066406, "learning_rate": 6.308283547151295e-08, "loss": 0.6588, "step": 3746 }, { "epoch": 2.85208452312964, "grad_norm": 29.01205825805664, "learning_rate": 6.244535910258697e-08, "loss": 0.5929, "step": 3747 }, { "epoch": 2.8528459927660386, "grad_norm": 40.740234375, "learning_rate": 6.18110998441257e-08, "loss": 0.7371, "step": 3748 }, { "epoch": 2.8536074624024366, "grad_norm": 38.27766799926758, "learning_rate": 6.118005810938687e-08, "loss": 0.8176, "step": 3749 }, { "epoch": 2.854368932038835, "grad_norm": 34.832435607910156, "learning_rate": 6.055223430953105e-08, "loss": 0.5896, "step": 3750 }, { "epoch": 2.8551304016752335, "grad_norm": 34.04582977294922, "learning_rate": 5.992762885362324e-08, "loss": 0.7374, "step": 3751 }, { "epoch": 2.8558918713116315, "grad_norm": 39.311641693115234, "learning_rate": 5.930624214863123e-08, "loss": 0.6748, "step": 3752 }, { "epoch": 2.8566533409480295, "grad_norm": 40.98863220214844, "learning_rate": 5.8688074599425046e-08, "loss": 0.9818, "step": 3753 }, { "epoch": 2.857414810584428, "grad_norm": 31.21957778930664, "learning_rate": 5.8073126608778064e-08, "loss": 0.7032, "step": 3754 }, { "epoch": 2.8581762802208264, "grad_norm": 37.621612548828125, "learning_rate": 5.7461398577364215e-08, "loss": 0.7556, "step": 3755 }, { "epoch": 2.8589377498572244, "grad_norm": 23.619646072387695, "learning_rate": 5.685289090376245e-08, "loss": 0.5245, "step": 3756 }, { "epoch": 2.859699219493623, "grad_norm": 33.746517181396484, "learning_rate": 5.6247603984450614e-08, "loss": 0.7809, "step": 3757 }, { "epoch": 2.860460689130021, "grad_norm": 41.407222747802734, "learning_rate": 5.564553821380936e-08, "loss": 0.7277, "step": 3758 }, { "epoch": 2.8612221587664193, "grad_norm": 40.51264953613281, "learning_rate": 5.5046693984120433e-08, "loss": 0.862, "step": 3759 }, { "epoch": 2.8619836284028173, "grad_norm": 36.89873504638672, "learning_rate": 5.445107168556618e-08, "loss": 0.687, "step": 3760 }, { "epoch": 2.8627450980392157, "grad_norm": 32.98456954956055, "learning_rate": 5.3858671706230605e-08, "loss": 0.6938, "step": 3761 }, { "epoch": 2.863506567675614, "grad_norm": 35.93764877319336, "learning_rate": 5.3269494432097166e-08, "loss": 0.7155, "step": 3762 }, { "epoch": 2.864268037312012, "grad_norm": 31.423114776611328, "learning_rate": 5.268354024704936e-08, "loss": 0.6355, "step": 3763 }, { "epoch": 2.86502950694841, "grad_norm": 42.10991668701172, "learning_rate": 5.210080953287122e-08, "loss": 0.9449, "step": 3764 }, { "epoch": 2.8657909765848086, "grad_norm": 51.16061019897461, "learning_rate": 5.1521302669247374e-08, "loss": 0.9807, "step": 3765 }, { "epoch": 2.866552446221207, "grad_norm": 51.38850402832031, "learning_rate": 5.0945020033760785e-08, "loss": 0.9054, "step": 3766 }, { "epoch": 2.867313915857605, "grad_norm": 37.3819694519043, "learning_rate": 5.0371962001893314e-08, "loss": 0.7033, "step": 3767 }, { "epoch": 2.8680753854940035, "grad_norm": 37.25703811645508, "learning_rate": 4.98021289470263e-08, "loss": 0.8228, "step": 3768 }, { "epoch": 2.8688368551304015, "grad_norm": 37.200897216796875, "learning_rate": 4.923552124044051e-08, "loss": 0.8252, "step": 3769 }, { "epoch": 2.8695983247668, "grad_norm": 37.42913055419922, "learning_rate": 4.8672139251313976e-08, "loss": 0.6444, "step": 3770 }, { "epoch": 2.870359794403198, "grad_norm": 38.48983383178711, "learning_rate": 4.811198334672418e-08, "loss": 0.8876, "step": 3771 }, { "epoch": 2.8711212640395964, "grad_norm": 38.742923736572266, "learning_rate": 4.755505389164472e-08, "loss": 0.7561, "step": 3772 }, { "epoch": 2.871882733675995, "grad_norm": 39.99129104614258, "learning_rate": 4.700135124894978e-08, "loss": 0.7405, "step": 3773 }, { "epoch": 2.872644203312393, "grad_norm": 32.57969665527344, "learning_rate": 4.64508757794091e-08, "loss": 0.5876, "step": 3774 }, { "epoch": 2.873405672948791, "grad_norm": 30.872175216674805, "learning_rate": 4.590362784169022e-08, "loss": 0.7185, "step": 3775 }, { "epoch": 2.8741671425851893, "grad_norm": 42.15765380859375, "learning_rate": 4.535960779235682e-08, "loss": 0.811, "step": 3776 }, { "epoch": 2.874928612221588, "grad_norm": 32.28451156616211, "learning_rate": 4.481881598587146e-08, "loss": 0.6001, "step": 3777 }, { "epoch": 2.875690081857986, "grad_norm": 26.834087371826172, "learning_rate": 4.4281252774591724e-08, "loss": 0.5702, "step": 3778 }, { "epoch": 2.8764515514943843, "grad_norm": 33.39504623413086, "learning_rate": 4.374691850877188e-08, "loss": 0.7696, "step": 3779 }, { "epoch": 2.8772130211307823, "grad_norm": 32.79410171508789, "learning_rate": 4.321581353656234e-08, "loss": 0.6475, "step": 3780 }, { "epoch": 2.8779744907671807, "grad_norm": 34.42647171020508, "learning_rate": 4.268793820401018e-08, "loss": 0.5369, "step": 3781 }, { "epoch": 2.8787359604035787, "grad_norm": 42.515655517578125, "learning_rate": 4.2163292855056936e-08, "loss": 0.8494, "step": 3782 }, { "epoch": 2.879497430039977, "grad_norm": 34.2271728515625, "learning_rate": 4.1641877831540834e-08, "loss": 0.6314, "step": 3783 }, { "epoch": 2.8802588996763756, "grad_norm": 32.17212677001953, "learning_rate": 4.1123693473194006e-08, "loss": 0.5725, "step": 3784 }, { "epoch": 2.8810203693127736, "grad_norm": 29.81683349609375, "learning_rate": 4.060874011764526e-08, "loss": 0.6275, "step": 3785 }, { "epoch": 2.881781838949172, "grad_norm": 30.778064727783203, "learning_rate": 4.009701810041677e-08, "loss": 0.5995, "step": 3786 }, { "epoch": 2.88254330858557, "grad_norm": 34.32253646850586, "learning_rate": 3.9588527754926276e-08, "loss": 0.7265, "step": 3787 }, { "epoch": 2.8833047782219685, "grad_norm": 32.055484771728516, "learning_rate": 3.908326941248486e-08, "loss": 0.4971, "step": 3788 }, { "epoch": 2.8840662478583665, "grad_norm": 38.89060592651367, "learning_rate": 3.858124340229863e-08, "loss": 0.5648, "step": 3789 }, { "epoch": 2.884827717494765, "grad_norm": 30.70846939086914, "learning_rate": 3.80824500514676e-08, "loss": 0.6622, "step": 3790 }, { "epoch": 2.8855891871311634, "grad_norm": 37.22810745239258, "learning_rate": 3.758688968498514e-08, "loss": 0.6014, "step": 3791 }, { "epoch": 2.8863506567675614, "grad_norm": 35.7891960144043, "learning_rate": 3.709456262573852e-08, "loss": 0.5603, "step": 3792 }, { "epoch": 2.8871121264039594, "grad_norm": 35.70576858520508, "learning_rate": 3.660546919450725e-08, "loss": 0.7021, "step": 3793 }, { "epoch": 2.887873596040358, "grad_norm": 36.806941986083984, "learning_rate": 3.611960970996642e-08, "loss": 0.6748, "step": 3794 }, { "epoch": 2.8886350656767563, "grad_norm": 33.336509704589844, "learning_rate": 3.563698448868058e-08, "loss": 0.6613, "step": 3795 }, { "epoch": 2.8893965353131543, "grad_norm": 40.31378173828125, "learning_rate": 3.515759384510986e-08, "loss": 0.8131, "step": 3796 }, { "epoch": 2.8901580049495528, "grad_norm": 39.15717315673828, "learning_rate": 3.4681438091605513e-08, "loss": 0.775, "step": 3797 }, { "epoch": 2.8909194745859508, "grad_norm": 31.80278205871582, "learning_rate": 3.4208517538411034e-08, "loss": 0.5008, "step": 3798 }, { "epoch": 2.8916809442223492, "grad_norm": 36.38859176635742, "learning_rate": 3.373883249366328e-08, "loss": 0.5307, "step": 3799 }, { "epoch": 2.8924424138587472, "grad_norm": 40.69465255737305, "learning_rate": 3.327238326338855e-08, "loss": 0.9228, "step": 3800 }, { "epoch": 2.8932038834951457, "grad_norm": 34.133697509765625, "learning_rate": 3.280917015150764e-08, "loss": 0.6896, "step": 3801 }, { "epoch": 2.893965353131544, "grad_norm": 58.15951919555664, "learning_rate": 3.2349193459829675e-08, "loss": 0.9049, "step": 3802 }, { "epoch": 2.894726822767942, "grad_norm": 34.269168853759766, "learning_rate": 3.1892453488058803e-08, "loss": 0.5859, "step": 3803 }, { "epoch": 2.89548829240434, "grad_norm": 40.491119384765625, "learning_rate": 3.143895053378698e-08, "loss": 0.768, "step": 3804 }, { "epoch": 2.8962497620407386, "grad_norm": 27.66516876220703, "learning_rate": 3.0988684892498955e-08, "loss": 0.6943, "step": 3805 }, { "epoch": 2.897011231677137, "grad_norm": 39.12172317504883, "learning_rate": 3.0541656857569494e-08, "loss": 0.6889, "step": 3806 }, { "epoch": 2.897772701313535, "grad_norm": 32.86622619628906, "learning_rate": 3.009786672026338e-08, "loss": 0.582, "step": 3807 }, { "epoch": 2.8985341709499335, "grad_norm": 35.131656646728516, "learning_rate": 2.96573147697371e-08, "loss": 0.8069, "step": 3808 }, { "epoch": 2.8992956405863315, "grad_norm": 40.5977897644043, "learning_rate": 2.9220001293036038e-08, "loss": 0.8461, "step": 3809 }, { "epoch": 2.90005711022273, "grad_norm": 37.44358825683594, "learning_rate": 2.87859265750956e-08, "loss": 0.7818, "step": 3810 }, { "epoch": 2.900818579859128, "grad_norm": 31.228519439697266, "learning_rate": 2.8355090898741778e-08, "loss": 0.6089, "step": 3811 }, { "epoch": 2.9015800494955264, "grad_norm": 26.14179801940918, "learning_rate": 2.7927494544690014e-08, "loss": 0.6281, "step": 3812 }, { "epoch": 2.902341519131925, "grad_norm": 30.46190071105957, "learning_rate": 2.7503137791543566e-08, "loss": 0.722, "step": 3813 }, { "epoch": 2.903102988768323, "grad_norm": 44.32353591918945, "learning_rate": 2.7082020915796813e-08, "loss": 0.8499, "step": 3814 }, { "epoch": 2.903864458404721, "grad_norm": 35.035465240478516, "learning_rate": 2.6664144191833053e-08, "loss": 0.695, "step": 3815 }, { "epoch": 2.9046259280411193, "grad_norm": 41.01001739501953, "learning_rate": 2.6249507891922264e-08, "loss": 0.791, "step": 3816 }, { "epoch": 2.9053873976775177, "grad_norm": 42.28317642211914, "learning_rate": 2.5838112286226123e-08, "loss": 0.9556, "step": 3817 }, { "epoch": 2.9061488673139158, "grad_norm": 34.69917297363281, "learning_rate": 2.5429957642792435e-08, "loss": 0.6865, "step": 3818 }, { "epoch": 2.906910336950314, "grad_norm": 31.544666290283203, "learning_rate": 2.502504422755847e-08, "loss": 0.7502, "step": 3819 }, { "epoch": 2.907671806586712, "grad_norm": 33.79249954223633, "learning_rate": 2.4623372304349858e-08, "loss": 0.6592, "step": 3820 }, { "epoch": 2.9084332762231107, "grad_norm": 34.30724334716797, "learning_rate": 2.4224942134878914e-08, "loss": 0.7457, "step": 3821 }, { "epoch": 2.9091947458595087, "grad_norm": 35.42524337768555, "learning_rate": 2.3829753978747427e-08, "loss": 0.8444, "step": 3822 }, { "epoch": 2.909956215495907, "grad_norm": 38.1048583984375, "learning_rate": 2.3437808093443314e-08, "loss": 0.8105, "step": 3823 }, { "epoch": 2.9107176851323056, "grad_norm": 38.365013122558594, "learning_rate": 2.304910473434341e-08, "loss": 1.0075, "step": 3824 }, { "epoch": 2.9114791547687036, "grad_norm": 36.25634765625, "learning_rate": 2.2663644154710117e-08, "loss": 0.7873, "step": 3825 }, { "epoch": 2.912240624405102, "grad_norm": 35.131717681884766, "learning_rate": 2.2281426605694212e-08, "loss": 0.7271, "step": 3826 }, { "epoch": 2.9130020940415, "grad_norm": 39.136924743652344, "learning_rate": 2.1902452336333147e-08, "loss": 0.6387, "step": 3827 }, { "epoch": 2.9137635636778985, "grad_norm": 38.332000732421875, "learning_rate": 2.1526721593551626e-08, "loss": 0.6601, "step": 3828 }, { "epoch": 2.9145250333142965, "grad_norm": 30.712629318237305, "learning_rate": 2.1154234622160487e-08, "loss": 0.6912, "step": 3829 }, { "epoch": 2.915286502950695, "grad_norm": 41.73023986816406, "learning_rate": 2.0784991664856702e-08, "loss": 0.6515, "step": 3830 }, { "epoch": 2.9160479725870934, "grad_norm": 33.604705810546875, "learning_rate": 2.0418992962224495e-08, "loss": 0.6349, "step": 3831 }, { "epoch": 2.9168094422234914, "grad_norm": 26.558271408081055, "learning_rate": 2.005623875273366e-08, "loss": 0.5122, "step": 3832 }, { "epoch": 2.9175709118598894, "grad_norm": 34.585323333740234, "learning_rate": 1.9696729272739577e-08, "loss": 0.7046, "step": 3833 }, { "epoch": 2.918332381496288, "grad_norm": 23.879518508911133, "learning_rate": 1.934046475648488e-08, "loss": 0.5155, "step": 3834 }, { "epoch": 2.9190938511326863, "grad_norm": 34.331695556640625, "learning_rate": 1.8987445436096653e-08, "loss": 0.6145, "step": 3835 }, { "epoch": 2.9198553207690843, "grad_norm": 28.932937622070312, "learning_rate": 1.8637671541587578e-08, "loss": 0.5891, "step": 3836 }, { "epoch": 2.9206167904054827, "grad_norm": 59.109493255615234, "learning_rate": 1.8291143300857018e-08, "loss": 0.917, "step": 3837 }, { "epoch": 2.9213782600418807, "grad_norm": 24.442825317382812, "learning_rate": 1.7947860939688255e-08, "loss": 0.4936, "step": 3838 }, { "epoch": 2.922139729678279, "grad_norm": 45.740821838378906, "learning_rate": 1.7607824681750152e-08, "loss": 0.5951, "step": 3839 }, { "epoch": 2.922901199314677, "grad_norm": 32.96456527709961, "learning_rate": 1.7271034748596592e-08, "loss": 0.606, "step": 3840 }, { "epoch": 2.9236626689510756, "grad_norm": 29.053844451904297, "learning_rate": 1.693749135966649e-08, "loss": 0.6389, "step": 3841 }, { "epoch": 2.924424138587474, "grad_norm": 33.63788986206055, "learning_rate": 1.660719473228267e-08, "loss": 0.6105, "step": 3842 }, { "epoch": 2.925185608223872, "grad_norm": 30.289072036743164, "learning_rate": 1.6280145081653543e-08, "loss": 0.6316, "step": 3843 }, { "epoch": 2.92594707786027, "grad_norm": 40.451393127441406, "learning_rate": 1.595634262087087e-08, "loss": 0.8085, "step": 3844 }, { "epoch": 2.9267085474966685, "grad_norm": 34.47736358642578, "learning_rate": 1.563578756091144e-08, "loss": 0.6642, "step": 3845 }, { "epoch": 2.927470017133067, "grad_norm": 39.166996002197266, "learning_rate": 1.5318480110636525e-08, "loss": 0.8558, "step": 3846 }, { "epoch": 2.928231486769465, "grad_norm": 34.527130126953125, "learning_rate": 1.5004420476790182e-08, "loss": 0.5917, "step": 3847 }, { "epoch": 2.9289929564058634, "grad_norm": 32.412193298339844, "learning_rate": 1.4693608864000953e-08, "loss": 0.69, "step": 3848 }, { "epoch": 2.9297544260422614, "grad_norm": 38.92666244506836, "learning_rate": 1.4386045474781285e-08, "loss": 0.7254, "step": 3849 }, { "epoch": 2.93051589567866, "grad_norm": 30.51943588256836, "learning_rate": 1.4081730509526992e-08, "loss": 0.6014, "step": 3850 }, { "epoch": 2.931277365315058, "grad_norm": 32.65416717529297, "learning_rate": 1.3780664166517244e-08, "loss": 0.6556, "step": 3851 }, { "epoch": 2.9320388349514563, "grad_norm": 38.0948371887207, "learning_rate": 1.3482846641914572e-08, "loss": 0.6216, "step": 3852 }, { "epoch": 2.932800304587855, "grad_norm": 31.22386932373047, "learning_rate": 1.3188278129764865e-08, "loss": 0.6958, "step": 3853 }, { "epoch": 2.933561774224253, "grad_norm": 44.911624908447266, "learning_rate": 1.2896958821997374e-08, "loss": 0.7515, "step": 3854 }, { "epoch": 2.934323243860651, "grad_norm": 36.4887809753418, "learning_rate": 1.2608888908424155e-08, "loss": 0.5772, "step": 3855 }, { "epoch": 2.9350847134970492, "grad_norm": 38.10952377319336, "learning_rate": 1.2324068576738401e-08, "loss": 0.7612, "step": 3856 }, { "epoch": 2.9358461831334477, "grad_norm": 41.66233444213867, "learning_rate": 1.2042498012518888e-08, "loss": 0.7793, "step": 3857 }, { "epoch": 2.9366076527698457, "grad_norm": 32.86172866821289, "learning_rate": 1.176417739922553e-08, "loss": 0.6234, "step": 3858 }, { "epoch": 2.937369122406244, "grad_norm": 42.22944259643555, "learning_rate": 1.1489106918200487e-08, "loss": 0.8396, "step": 3859 }, { "epoch": 2.938130592042642, "grad_norm": 30.516939163208008, "learning_rate": 1.1217286748668176e-08, "loss": 0.6248, "step": 3860 }, { "epoch": 2.9388920616790406, "grad_norm": 30.446134567260742, "learning_rate": 1.0948717067735815e-08, "loss": 0.5613, "step": 3861 }, { "epoch": 2.9396535313154386, "grad_norm": 34.16997528076172, "learning_rate": 1.0683398050392869e-08, "loss": 0.5438, "step": 3862 }, { "epoch": 2.940415000951837, "grad_norm": 34.66270065307617, "learning_rate": 1.0421329869509944e-08, "loss": 0.7424, "step": 3863 }, { "epoch": 2.9411764705882355, "grad_norm": 29.576295852661133, "learning_rate": 1.0162512695840454e-08, "loss": 0.583, "step": 3864 }, { "epoch": 2.9419379402246335, "grad_norm": 29.63631248474121, "learning_rate": 9.906946698018949e-09, "loss": 0.5368, "step": 3865 }, { "epoch": 2.942699409861032, "grad_norm": 36.934329986572266, "learning_rate": 9.654632042562229e-09, "loss": 0.7233, "step": 3866 }, { "epoch": 2.94346087949743, "grad_norm": 27.11764907836914, "learning_rate": 9.405568893868232e-09, "loss": 0.4906, "step": 3867 }, { "epoch": 2.9442223491338284, "grad_norm": 33.39051055908203, "learning_rate": 9.159757414215487e-09, "loss": 0.7552, "step": 3868 }, { "epoch": 2.9449838187702264, "grad_norm": 31.640033721923828, "learning_rate": 8.917197763765873e-09, "loss": 0.6498, "step": 3869 }, { "epoch": 2.945745288406625, "grad_norm": 31.906227111816406, "learning_rate": 8.677890100561303e-09, "loss": 0.5181, "step": 3870 }, { "epoch": 2.9465067580430233, "grad_norm": 41.91764831542969, "learning_rate": 8.441834580524278e-09, "loss": 0.7096, "step": 3871 }, { "epoch": 2.9472682276794213, "grad_norm": 28.97077751159668, "learning_rate": 8.20903135745954e-09, "loss": 0.6077, "step": 3872 }, { "epoch": 2.9480296973158193, "grad_norm": 40.00917053222656, "learning_rate": 7.979480583052423e-09, "loss": 0.7153, "step": 3873 }, { "epoch": 2.9487911669522178, "grad_norm": 37.50101089477539, "learning_rate": 7.753182406868288e-09, "loss": 0.8325, "step": 3874 }, { "epoch": 2.949552636588616, "grad_norm": 26.70798110961914, "learning_rate": 7.530136976354186e-09, "loss": 0.5173, "step": 3875 }, { "epoch": 2.950314106225014, "grad_norm": 35.37315368652344, "learning_rate": 7.310344436837758e-09, "loss": 0.7255, "step": 3876 }, { "epoch": 2.9510755758614127, "grad_norm": 37.36168670654297, "learning_rate": 7.093804931526116e-09, "loss": 0.7439, "step": 3877 }, { "epoch": 2.9518370454978107, "grad_norm": 29.593427658081055, "learning_rate": 6.880518601508624e-09, "loss": 0.6742, "step": 3878 }, { "epoch": 2.952598515134209, "grad_norm": 39.19407653808594, "learning_rate": 6.670485585753006e-09, "loss": 0.7168, "step": 3879 }, { "epoch": 2.953359984770607, "grad_norm": 36.97806930541992, "learning_rate": 6.4637060211092395e-09, "loss": 0.8665, "step": 3880 }, { "epoch": 2.9541214544070056, "grad_norm": 22.17021942138672, "learning_rate": 6.260180042305663e-09, "loss": 0.429, "step": 3881 }, { "epoch": 2.954882924043404, "grad_norm": 30.870136260986328, "learning_rate": 6.0599077819523126e-09, "loss": 0.6389, "step": 3882 }, { "epoch": 2.955644393679802, "grad_norm": 35.432186126708984, "learning_rate": 5.862889370538138e-09, "loss": 0.7674, "step": 3883 }, { "epoch": 2.9564058633162, "grad_norm": 40.02390670776367, "learning_rate": 5.669124936432124e-09, "loss": 0.8454, "step": 3884 }, { "epoch": 2.9571673329525985, "grad_norm": 37.33639907836914, "learning_rate": 5.478614605883281e-09, "loss": 0.7239, "step": 3885 }, { "epoch": 2.957928802588997, "grad_norm": 35.70606994628906, "learning_rate": 5.2913585030206495e-09, "loss": 0.6957, "step": 3886 }, { "epoch": 2.958690272225395, "grad_norm": 33.54386901855469, "learning_rate": 5.107356749853298e-09, "loss": 0.5057, "step": 3887 }, { "epoch": 2.9594517418617934, "grad_norm": 48.0322265625, "learning_rate": 4.926609466268106e-09, "loss": 0.7479, "step": 3888 }, { "epoch": 2.9602132114981914, "grad_norm": 35.88151931762695, "learning_rate": 4.749116770033646e-09, "loss": 0.6382, "step": 3889 }, { "epoch": 2.96097468113459, "grad_norm": 35.41067886352539, "learning_rate": 4.574878776796299e-09, "loss": 0.7424, "step": 3890 }, { "epoch": 2.961736150770988, "grad_norm": 38.299678802490234, "learning_rate": 4.40389560008303e-09, "loss": 0.6551, "step": 3891 }, { "epoch": 2.9624976204073863, "grad_norm": 33.10376739501953, "learning_rate": 4.236167351299169e-09, "loss": 0.6503, "step": 3892 }, { "epoch": 2.9632590900437847, "grad_norm": 31.249210357666016, "learning_rate": 4.0716941397300755e-09, "loss": 0.6344, "step": 3893 }, { "epoch": 2.9640205596801827, "grad_norm": 38.98311233520508, "learning_rate": 3.910476072539471e-09, "loss": 0.6182, "step": 3894 }, { "epoch": 2.9647820293165807, "grad_norm": 30.86043930053711, "learning_rate": 3.752513254770551e-09, "loss": 0.6764, "step": 3895 }, { "epoch": 2.965543498952979, "grad_norm": 45.20832824707031, "learning_rate": 3.597805789344877e-09, "loss": 0.7468, "step": 3896 }, { "epoch": 2.9663049685893776, "grad_norm": 45.20635223388672, "learning_rate": 3.446353777065148e-09, "loss": 0.8212, "step": 3897 }, { "epoch": 2.9670664382257756, "grad_norm": 39.95635986328125, "learning_rate": 3.2981573166096515e-09, "loss": 0.885, "step": 3898 }, { "epoch": 2.967827907862174, "grad_norm": 34.16203689575195, "learning_rate": 3.15321650453837e-09, "loss": 0.5811, "step": 3899 }, { "epoch": 2.968589377498572, "grad_norm": 37.86811447143555, "learning_rate": 3.0115314352885393e-09, "loss": 0.8093, "step": 3900 }, { "epoch": 2.9693508471349706, "grad_norm": 38.76040267944336, "learning_rate": 2.8731022011757593e-09, "loss": 0.6834, "step": 3901 }, { "epoch": 2.9701123167713686, "grad_norm": 38.24364471435547, "learning_rate": 2.7379288923962133e-09, "loss": 0.758, "step": 3902 }, { "epoch": 2.970873786407767, "grad_norm": 30.111642837524414, "learning_rate": 2.6060115970216736e-09, "loss": 0.7318, "step": 3903 }, { "epoch": 2.9716352560441655, "grad_norm": 37.600894927978516, "learning_rate": 2.4773504010056072e-09, "loss": 0.8749, "step": 3904 }, { "epoch": 2.9723967256805635, "grad_norm": 34.56288146972656, "learning_rate": 2.3519453881781784e-09, "loss": 0.6759, "step": 3905 }, { "epoch": 2.9731581953169615, "grad_norm": 31.178369522094727, "learning_rate": 2.2297966402473616e-09, "loss": 0.5763, "step": 3906 }, { "epoch": 2.97391966495336, "grad_norm": 52.18911361694336, "learning_rate": 2.110904236800604e-09, "loss": 0.8106, "step": 3907 }, { "epoch": 2.9746811345897584, "grad_norm": 41.237213134765625, "learning_rate": 1.9952682553042722e-09, "loss": 0.9387, "step": 3908 }, { "epoch": 2.9754426042261564, "grad_norm": 37.38624954223633, "learning_rate": 1.882888771100877e-09, "loss": 0.8985, "step": 3909 }, { "epoch": 2.976204073862555, "grad_norm": 36.6756477355957, "learning_rate": 1.7737658574135118e-09, "loss": 0.581, "step": 3910 }, { "epoch": 2.976965543498953, "grad_norm": 34.74436569213867, "learning_rate": 1.6678995853408597e-09, "loss": 0.747, "step": 3911 }, { "epoch": 2.9777270131353513, "grad_norm": 36.89299392700195, "learning_rate": 1.565290023862187e-09, "loss": 0.9182, "step": 3912 }, { "epoch": 2.9784884827717493, "grad_norm": 37.64173889160156, "learning_rate": 1.4659372398334593e-09, "loss": 0.7529, "step": 3913 }, { "epoch": 2.9792499524081477, "grad_norm": 31.766340255737305, "learning_rate": 1.3698412979890051e-09, "loss": 0.6699, "step": 3914 }, { "epoch": 2.980011422044546, "grad_norm": 32.65924072265625, "learning_rate": 1.2770022609409628e-09, "loss": 0.6034, "step": 3915 }, { "epoch": 2.980772891680944, "grad_norm": 38.42402267456055, "learning_rate": 1.187420189178723e-09, "loss": 0.8328, "step": 3916 }, { "epoch": 2.9815343613173426, "grad_norm": 41.06101608276367, "learning_rate": 1.1010951410717063e-09, "loss": 0.7392, "step": 3917 }, { "epoch": 2.9822958309537406, "grad_norm": 30.401086807250977, "learning_rate": 1.0180271728649215e-09, "loss": 0.5825, "step": 3918 }, { "epoch": 2.983057300590139, "grad_norm": 35.010921478271484, "learning_rate": 9.382163386822962e-10, "loss": 0.776, "step": 3919 }, { "epoch": 2.983818770226537, "grad_norm": 31.228429794311523, "learning_rate": 8.616626905255665e-10, "loss": 0.5924, "step": 3920 }, { "epoch": 2.9845802398629355, "grad_norm": 30.267112731933594, "learning_rate": 7.883662782737223e-10, "loss": 0.6111, "step": 3921 }, { "epoch": 2.985341709499334, "grad_norm": 35.103424072265625, "learning_rate": 7.18327149683562e-10, "loss": 0.767, "step": 3922 }, { "epoch": 2.986103179135732, "grad_norm": 41.2313346862793, "learning_rate": 6.515453503896929e-10, "loss": 0.7436, "step": 3923 }, { "epoch": 2.98686464877213, "grad_norm": 31.063430786132812, "learning_rate": 5.88020923905086e-10, "loss": 0.5943, "step": 3924 }, { "epoch": 2.9876261184085284, "grad_norm": 44.756988525390625, "learning_rate": 5.277539116194108e-10, "loss": 0.7221, "step": 3925 }, { "epoch": 2.988387588044927, "grad_norm": 40.38260269165039, "learning_rate": 4.707443527995903e-10, "loss": 0.815, "step": 3926 }, { "epoch": 2.989149057681325, "grad_norm": 37.2851448059082, "learning_rate": 4.1699228459202156e-10, "loss": 0.6248, "step": 3927 }, { "epoch": 2.9899105273177233, "grad_norm": 35.07923889160156, "learning_rate": 3.6649774201813483e-10, "loss": 0.7911, "step": 3928 }, { "epoch": 2.9906719969541213, "grad_norm": 32.70376205444336, "learning_rate": 3.1926075797827914e-10, "loss": 0.6661, "step": 3929 }, { "epoch": 2.99143346659052, "grad_norm": 31.337936401367188, "learning_rate": 2.752813632506124e-10, "loss": 0.6612, "step": 3930 }, { "epoch": 2.992194936226918, "grad_norm": 31.936630249023438, "learning_rate": 2.345595864905459e-10, "loss": 0.6898, "step": 3931 }, { "epoch": 2.9929564058633162, "grad_norm": 40.11835861206055, "learning_rate": 1.970954542301895e-10, "loss": 0.7675, "step": 3932 }, { "epoch": 2.9937178754997147, "grad_norm": 38.77583694458008, "learning_rate": 1.6288899088001687e-10, "loss": 0.838, "step": 3933 }, { "epoch": 2.9944793451361127, "grad_norm": 30.902219772338867, "learning_rate": 1.319402187272001e-10, "loss": 0.6359, "step": 3934 }, { "epoch": 2.9952408147725107, "grad_norm": 42.365943908691406, "learning_rate": 1.0424915793671997e-10, "loss": 0.8253, "step": 3935 }, { "epoch": 2.996002284408909, "grad_norm": 48.328121185302734, "learning_rate": 7.9815826551366e-11, "loss": 0.7837, "step": 3936 }, { "epoch": 2.9967637540453076, "grad_norm": 37.16790771484375, "learning_rate": 5.864024049062612e-11, "loss": 0.6757, "step": 3937 }, { "epoch": 2.9975252236817056, "grad_norm": 30.08350944519043, "learning_rate": 4.0722413551241845e-11, "loss": 0.6055, "step": 3938 }, { "epoch": 2.998286693318104, "grad_norm": 30.512596130371094, "learning_rate": 2.606235740887364e-11, "loss": 0.5374, "step": 3939 }, { "epoch": 2.999048162954502, "grad_norm": 32.8347282409668, "learning_rate": 1.466008161421506e-11, "loss": 0.556, "step": 3940 }, { "epoch": 2.9998096325909005, "grad_norm": 35.26900863647461, "learning_rate": 6.5155935974337e-12, "loss": 0.6257, "step": 3941 }, { "epoch": 3.0, "grad_norm": 79.82987976074219, "learning_rate": 1.6288986648405058e-12, "loss": 0.7447, "step": 3942 } ], "logging_steps": 1, "max_steps": 3942, "num_input_tokens_seen": 0, "num_train_epochs": 3, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 2.1640587210986342e+17, "train_batch_size": 2, "trial_name": null, "trial_params": null }